Synthesis, biological activity and structure-activity relationships of new benzoic acid-based protein tyrosine phosphatase inhibitors endowed with insulinomimetic effects in mouse C2C12 skeletal muscle cells

Article abstract of DOI:10.1016/j.ejmech.2013.11.001

Insulin resistance is a complex altered metabolic condition characterized by impaired insulin signaling and implicated in the pathogenesis of serious human diseases, such as diabetes, obesity, neurodegenerative pathologies. In pursuing our aim to identify new agents able to improve cellular insulin sensitivity, we have synthesized new 4-[(5-arylidene-4-oxo-2-phenylimino/ oxothiazolidin-3-yl)methyl]benzoic acids (5, 8) and evaluated their inhibitory activity towards human protein tyrosine phosphatases PTP1B, LMW-PTP and TCPTP, enzymes which are involved in the development of insulin resistance. Compounds 5 and 8 showed from moderate to significant selectivity toward PTP1B over both the highly homologous TCPTP and the two isoforms of human LMW-PTP. In addition, most of the tested compounds selectively inhibited LMW-PTP IF1 over the isoform IF2. Docking studies into the active sites of PTP1B and LMW-PTP aided the rationalization of the observed PTP inhibitory profile. Moreover, most tested compounds were capable to induce the insulin metabolic pathway in mouse C2C12 skeletal muscle cells by remarkably stimulating both IRβ phosphorylation and 2-deoxyglucose cellular uptake.

Full text of DOI:10.1016/j.ejmech.2013.11.001

European Journal of Medicinal Chemistry 71 (2014) 112e127
Contents lists available at ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
Original article
Synthesis, biological activity and structureeactivity relationships of
new benzoic acid-based protein tyrosine phosphatase inhibitors
endowed with insulinomimetic effects in mouse C2C12 skeletal
muscle cells
,
b,
Rosaria Ottanà a, Rosanna Maccari a , Jérémie Mortier c, Anna Caselli d, Simona Amuso a,
*
Guido Camici d e, Archimede Rotondo f, Gerhard Wolber b, Paolo Paoli d
,
a Dipartimento di Scienze del Farmaco e dei Prodotti per la Salute, University of Messina, Polo Universitario dell’Annunziata, Viale SS. Annunziata,
98168 Messina, Italy
b Institute of Pharmacy, Department Pharmaceutical & Medicinal Chemistry, Freie Universität Berlin, 14195 Berlin, Germany
c Institute of Chemistry and Biochemistry, Organic Chemistry, Freie University Berlin, 14195 Berlin, Germany
d Dipartimento di Scienze Biomediche Sperimentali e Cliniche, Sezione di Scienze Biochimiche, University of Firenze, Viale Morgagni 50, 50134 Firenze, Italy
e Center for Research, Transfer and High Education DENOthe, Viale Pieraccini 6, 50139 Firenze, Italy
f Dipartimento di Scienze Chimiche, University of Messina, Viale F. Stagno d’Alcontres 31, 98166 Messina, Italy
a r t i c l e i n f o
a b s t r a c t
Article history:
Insulin resistance is a complex altered metabolic condition characterized by impaired insulin signaling
and implicated in the pathogenesis of serious human diseases, such as diabetes, obesity, neurodegen-
erative pathologies. In pursuing our aim to identify new agents able to improve cellular insulin sensi-
tivity, we have synthesized new 4-[(5-arylidene-4-oxo-2-phenylimino/oxothiazolidin-3-yl)methyl]
benzoic acids (5, 8) and evaluated their inhibitory activity towards human protein tyrosine phosphatases
PTP1B, LMW-PTP and TCPTP, enzymes which are involved in the development of insulin resistance.
Compounds 5 and 8 showed from moderate to significant selectivity toward PTP1B over both the highly
homologous TCPTP and the two isoforms of human LMW-PTP. In addition, most of the tested compounds
selectively inhibited LMW-PTP IF1 over the isoform IF2. Docking studies into the active sites of PTP1B and
LMW-PTP aided the rationalization of the observed PTP inhibitory profile. Moreover, most tested com-
pounds were capable to induce the insulin metabolic pathway in mouse C2C12 skeletal muscle cells by
Received 25 July 2013
Received in revised form
24 October 2013
Accepted 1 November 2013
Available online 11 November 2013
Keywords:
Protein tyrosine phosphatases
Enzyme inhibitors
Insulinomimetic effects
Molecular docking
5-Arylidene-4-thiazolidinone derivatives
4-[(5-Arylidene-4-oxo-2-phenylimino/
oxothiazolidin-3-yl)methyl]benzoic acids
remarkably stimulating both IR
b
phosphorylation and 2-deoxyglucose cellular uptake.
Ó 2013 Elsevier Masson SAS. All rights reserved.
1. Introduction
population by 2030. More than 90% of diabetic patients suffer from
type 2 diabetes (T2DM) [1]. The onset of T2DM is mainly linked to
the development of insulin resistance, an altered metabolic con-
dition in which the response of target tissues (such as fat tissue,
liver and skeletal muscle) to the hormone is significantly impaired.
Attenuated insulin signaling results in numerous metabolic and
cellular alterations, including diminished cellular glucose uptake,
increased liver gluconeogenesis, hyperglycemia and hyperlipid-
emia. Increased levels of blood glucose and lipids can induce
further deterioration of both insulin action and secretion as well as
trigger sequences of tissue and vascular dysfunctions that lead to
chronic diabetic complications and increased cardiovascular risk
[2,3].
Diabetes mellitus is a serious metabolic disease, which currently
affects more than 371 million people worldwide, with a global
prevalence of 8.3%. The number of people with diabetes is rapidly
growing in every country, correspondingly with the current obesity
epidemic, and is predicted to rise to almost 10% of total world
Abbreviations: 2-DOG, 2-deoxyglucose; IR, insulin receptor; LMW-PTP, low
molecular weight protein tyrosine phosphatase; MALDI-TOF MS, matrix-assisted
laser desorption/ionization time-of-flight mass; MTT, 3-(4,5-dimethylthiazol-2-yl)-
2,5-diphenyltetrazolium bromide; PTP, protein tyrosine phosphatase; STAT, signal
transducer and activator of transcription; TCPTP, T cell protein tyrosine phospha-
tase; T2DM, type 2 diabetes mellitus.
The multiple metabolic and tissue abnormalities induced by
insulin resistance can contribute to the development of metabolic
syndrome, a cluster of correlated comorbid disorders that, beside
* Corresponding author. Tel.: þ39 090 6766406.
E-mail address: rmaccari@unime.it (R. Maccari).
0223-5234/$ e see front matter Ó 2013 Elsevier Masson SAS. All rights reserved.
http://dx.doi.org/10.1016/j.ejmech.2013.11.001

R. Ottanà et al. / European Journal of Medicinal Chemistry 71 (2014) 112e127
113
T2DM or impaired glucose tolerance, include obesity, dyslipidemia,
hypertension, increased levels of inflammatory mediators, endo-
thelial dysfunction, impaired thrombolysis and increased risk of
cerebrovascular and cardiovascular diseases [2,4].
Obesity can contribute to further attenuate the sensitivity of
target tissues to insulin. In fact, a number of cytokines produced by
Dysfunctions in PTP1B activity or expression can be induced
by several factors, such as excessive release of free fatty acids,
TNFa and pro-inflammatory cytokines. It has been demonstrated
that PTP1B overexpression may lead to increased gene expression
of the sterol regulatory element-binding protein-1 (SREBP-1), a
transcriptional factor which plays a pivotal role in increasing
fatty acid synthesis and triglyceride accumulation in liver.
Through this mechanism, PTP1B mediates hepatic lipogenesis
and postprandial hypertriglyceridemia, which are thought to be
important factors in the pathogenesis of insulin resistance and
metabolic syndrome [19].
Distinct studies demonstrated that, in mice, PTP1B knock-down
enhances insulin sensitivity and resistance to diet-induced obesity,
without causing abnormalities in growth or other vital functions
[14,15]. Muscle-specific or liver-specific deletion of PTP1B led to
improved glucose homeostasis and enhanced systemic insulin
sensitivity in diet-induced insulin resistant animals, demonstrating
that this phosphatase plays a crucial tissue-specific role in the
control of insulin signaling, mainly in skeletal muscle and liver but
not in adipose tissue [15,20e22]. PTP1B deficiency in both myocyte
adipose tissue (adipokines), such as tumor necrosis factor-
a
(TNF
a)
and leptin, can influence the response to the hormone. TNF
a
im-
pairs several components of the insulin signal pathway, thus
causing reduced glucose uptake and insulin resistance of target
tissues [4,5]. Leptin activates specific receptors in hypothalamus,
thus reducing food intake and increasing energy expenditure, and
also exerts direct effects on cellular insulin sensitivity. States of
leptin deficiency or resistance, which are associated to obesity, lead
to lipid overaccumulation in nonadipose tissues, such as liver and
muscle, thus attenuating insulin signaling and contributing to the
development of insulin resistance [6,7].
Recent compelling evidence supports the pivotal roles of insulin
in neuronal survival, synaptic plasticity and memory processing
and highlights a direct implication of insulin resistance also in the
pathogenesis of neurodegenerative diseases, such as Alzheimer’s
disease (AD), particularly in agedness [8e10]. AD has even been
described as “type 3 diabetes” and it has been suggested that it
could be considered mainly as a metabolic disease characterized by
a progressive loss of the brain capacity to respond to insulin and
insulin-like growth factor stimulation that results in cognitive
impairment [11]. Peripheral insulin resistance contributes to the
development of brain insulin resistance by inducing the production
of toxic lipids that are able to promote neurodegeneration [11].
Thus, it has been suggested that the soaring incidence of diseases
characterized by peripheral insulin resistance, such as obesity,
T2DM and metabolic syndrome, might represent an important
factor contributing to the current AD epidemic.
Considering that insulin resistance is one of the most prevalent
metabolic disorders worldwide and is crucially implicated in the
pathogenesis of common serious chronic diseases, there is an ur-
gent need to discover new drugs able to treat this pathological
condition.
The actions of insulin are mediated by multiple signal pathways
triggered by the insulin receptor (IR) tyrosine kinase which acts on
different intracellular substrates [12]. In this context, the patho-
physiological events that underlie the development of insulin
resistance are not fully understood and, consequently, fighting this
condition may be a challenging task.
cultures and in mice confers protection against TNF
sulin resistance [23].
a-induced in-
Additionally, both systemic and neuron-specific PTP1B
knockout mice are lean and hypersensitive to leptin, whereas
specific PTP1B deletion in muscle or liver did not affect body
weight, indicating that PTP1B regulates body weight and adiposity
mainly through the central control of leptin signaling in hypo-
thalamus [6,18,24]. In insulin resistant animals, elevated levels of
PTP1B were also found in other specific brain regions, such as ce-
rebral cortex and hippocampus, suggesting a possible general role
of the enzyme in the impairment of brain insulin signaling [25].
The correlation of increased PTP1B expression with insulin
resistance has also been demonstrated in humans [16]. In clinical
trials, the administration of PTP1B antisense oligonucleotides (ASO)
increased insulin sensitivity without inducing hypoglycemia [26].
Therefore, PTP1B is a promising validated target for the man-
agement of insulin resistance and, on this basis, the development of
PTP1B inhibitors is considered a significant advancement in the
search for new agents to treat T2DM, obesity and related
pathologies.
Unfortunately, the search for potent, sufficiently selective and
orally available PTP1B inhibitors has encountered difficulties,
mainly due to the polar nature and high degree of homology of PTP
catalytic sites. Despite the wealth of compounds so far investigated,
only few inhibitors have been evaluated as drug candidates in
clinical trials [27e30]. Currently, trodusquemine (Fig. 1), a selective
peripheral and central PTP1B inhibitor, has reached phase 2 clinical
trials as a promising antiobesity and antidiabetic agent [31].
However, in the design of active-site directed inhibitors of PTPs,
targeting both catalytic site and unique loops or subpockets adja-
cent to the active site is a useful approach to achieve high-affinity
binding and, consequently, effective and possibly selective inhibi-
tion [27,32e34].
However, several alterations in post-receptor events of insulin
cascade that lead to attenuated transduction of the hormone
signal have been identified as major causes of insulin resistance in
target tissues [3,13]. Out of these molecular mechanisms, the
increased activity and/or overexpression of the cytosolic protein
tyrosine phosphatase 1B (PTP1B) have been recognized to be key
components in the development of insulin resistance, as estab-
lished by numerous studies in both animal models and humans
[14e17].
PTP1B acts as a crucial negative regulator of both insulin and
leptin signal transduction. It controls the response of target tissues
to insulin by inactivating IR through the dephosphorylation of
In the structure of PTP1B, the catalytic domain is situated at the
bottom of a deep crevice which is surrounded by both the flexible
WPD loop, containing the catalytic residue Asp181, and the YRD
loop, including Tyr46 which is critical for pTyr recognition. The
phosphate-binding loop (P-loop) contains the signature motif CX5R
(S/T) that is conserved in all members of the PTP family and, in fact,
includes the cysteine and arginine residues that are essential for the
catalytic dephosphorylation mechanism common to all PTPs. In
PTP1B, a secondary noncatalytic arylphosphate-binding site adja-
cent to the active site is also present, which is lined by poorly
conserved residues, such as Arg24, Ala27, Phe52, Arg254, Met258,
Gly259. This secondary pocket, which is lacking in several other
specific phosphotyrosine (pTyr) residues on the receptor b subunits
as well as by dephosphorylating insulin substrate proteins (IRS),
which are downstream components of the insulin signal cascade
[12]. Moreover, PTP1B negatively regulates leptin signaling by
dephosphorylating and inactivating the Janus 2 kinase that is
associated with leptin receptor, thus preventing the activation of
downstream molecule STAT3 (signal transducer and activator of
transcription 3) and critically modifying the action of leptin in the
regulation of food intake and body weight [6,18].

114
R. Ottanà et al. / European Journal of Medicinal Chemistry 71 (2014) 112e127
Fig. 1. a) Structure of trodusquemine. b) General formula of benzoic acid-based derivatives 1, 2 active as human PTP1B/LMW-PTP IF1 dual inhibitors and structures of selected
inhibitors of the series [Refs. [50,51,53]].
PTPs, represents an important structural feature of this phospha-
tase in order to design potent bidentate ligands that can effectively
fit both the active site and the nearby noncatalytic site pocket [35].
Besides PTP1B, other PTPs (LMW-PTP, SHP2, LAR-PTP, TC-PTP)
appear to be involved in the control of insulin signaling [36].
Among them, low molecular weight protein tyrosine phosphatase
(LMW-PTP) has been recognized to play a key role in the control of
both insulin-mediated mitotic and metabolic signaling pathways
in liver and adipose tissue and did not induce any compensatory
increase in the other PTPs involved in the control of insulin
response. In addition, the treatment of animals with specific LMW-
PTP ASO did not bring about any toxic effect and was well tolerated
[38].
Although the three-dimensional structure of LMW-PTP is rather
different from those of higher molecular weight PTPs, such as
PTP1B, its catalytic site can be structurally superimposed on that of
PTP1B. The edge of the catalytic pocket is formed by two opposite
loops, one of which contains residues Tyr131 and Tyr132 which are
important for the substrate recognition. The other wall of the LMW-
PTP active site cavity is formed by a loop which significantly differs
in the two active isoforms of the human enzyme, IF1 and IF2. This
variable loop contains different amino acid residues, particularly
those in positions 49, 50 and 53, that can determine isoform
by dephosphorylating specific pTyr residues on the IRb subunits
[37]. In insulin resistant obese mice, LMW-PTP gene silencing by
means of specific ASO improved cellular sensitivity to insulin,
resulting in reduced insulinemic and glycemic levels and increased
activity of downstream components of the insulin cascade,
without influencing animal body weight. Interestingly, LMW-PTP
knock-down improved cellular response to the hormone mainly

R. Ottanà et al. / European Journal of Medicinal Chemistry 71 (2014) 112e127
115
specificity in the binding to substrates and modulating ligands
[34,39,40].
T cell protein tyrosine phosphatase (TCPTP) is involved in the
regulation of insulin signal transduction, by dephosphorylating
to their 2,4-thiazolidinedione counterparts 1. Moreover, several 2-
arylimino substituted derivatives, such as compound 2d (Fig. 1),
exhibited excellent insulinomimetic effects in skeletal muscle cell
cultures [53].
critical pTyr residues of IR
b
subunit in a manner similar to PTP1B
Starting from the finding that the most active PTP1B/IF1 in-
hibitors of both series 1 and 2 were those with a phenoxy- or
benzyloxybenzylidene moiety at position 5 of the thiazolidinone
[41]. Human TCPTP and PTP1B are closely related enzymes, dis-
playing a high degree of sequence identity and similarity in the
catalytic domain [42]. Despite their similarity, these PTPs have
distinct biological functions and can target different cellular sub-
strates, playing cooperative nonredundant and temporally-distinct
roles in the regulation of IR phosphorylation and glucose homeo-
stasis [43e45].
Interestingly, increased levels of both TCPTP and PTP1B were
found to cooperate not only to insulin resistance but also to diet-
induced leptin resistance in mice [46]. TCPTP expression is
induced by hyperleptinemia that may accompany leptin resistance
in obesity and, in turn, the increased hypothalamic TCPTP levels
impair leptin signal transduction by means of the selective
dephosphorylation of STAT3 in the nucleus. Accordingly, defi-
ciency of both PTP1B and TCPTP in mice determined additive ef-
fects in enhancing leptin sensitivity and improving glucose
homeostasis [46].
While homozygous TCPTP deletion led to hematopoietic and
immune disorders and consequent lethality in mice, TCPTPþ/ꢀ
heterozygous mice showed normal phenotype and life span. On
this basis, it has been suggested that the non-selective partial in-
hibition of both PTP1B and TCPTP might be useful to enhance the
efficacy of dual inhibitors in improving insulin and leptin signal
transduction [47e49].
scaffold (Fig. 1), we have synthesized
a series of 4-{[5-(2-
phenylethoxy)benzylidene-4-oxothiazolidin-3-yl]methyl}benzoic
acids 5aed and 8aed (Fig. 2), as part of a systematic optimization of
this class of benzoic acid-based inhibitors. 5-Benzyloxybenzylidene
substituted analogs 5eeg and 8eeg (Fig. 2) were also synthesized
and included in this study.
The depth of the PTP1B noncatalytic pocket, which is wider and
more lipophilic compared to those of other PTPs, provides an op-
portunity to achieve a gain in both affinity and selectivity for the
enzyme. Therefore, the extension of the 5-arylidene moiety or the
introduction of suitable substituents in this portion may be useful
to fit this region of the target enzyme more efficaciously. The in-
crease of the distance between the two aromatic rings might be
favorable to efficiently reach hydrophobic amino acid residues that
are situated deeply in this site, such as Phe52 and Ala27, which are
not conserved in all PTPs.
Compounds 5 and 8 were tested in vitro as inhibitors toward
recombinant human PTP1B as well as toward human LMW-PTP
isoforms (IF1 and IF2) and TCPTP. Kinetic studies with PTP1B
were also performed. Compounds 5 and 8 were further screened to
assess their capability to enhance the IR tyrosine phosphorylation
and glucose uptake in cultures of mouse differentiated C2C12
myoblasts, which express both insulin receptor and PTP1B.
Docking simulations into the PTP1B and LMW-PTP active sites
were carried out to rationalize the results and observed SARs.
In the last few years, in pursuing our goal to discover novel
small molecule nonphosphorous PTP inhibitors, we designed and
investigated
4-[(5-arylidene-2,4-dioxothiazolidin-3-yl)methyl]
benzoic acids 1 and their 2-arylimino analogs 2 (Fig. 1) which
represent a new class of human PTP1B/LMW-PTP dual inhibitors
with a monocarboxylic acid-based pTyr-mimic [50e53]. According
to our design proposal, the inhibition of both these PTPs could be
an opportunity for a novel multitarget approach to the treatment
of insulin resistance and related pathologies. Compounds 1 and 2
showed a generally appreciable selectivity towards PTP1B and
LMW-PTP IF1 over the isoform IF2. This may be a promising
feature, since IF1 appears to be the only LMW-PTP isoenzyme
significantly correlated to the increase of blood glucose levels in
T2DM patients [54].
In agreement with the initial structure-based design hypothesis,
SAR and molecular docking studies supported the specific action of
this structural class of PTP inhibitors and allowed the identification
of the molecular portions of benzoic acid derivatives 1 and 2 that
are crucial to ensure optimal interactions with both PTP1B and
LMW-PTP. These are: i) the p-methylbenzoic acid residue in posi-
tion 3 of the thiazolidinone ring, which can function as an efficient
nonphosphorous monoanionic pTyr-mimic targeting the catalytic
site; and ii) the 5-arylidene moiety, which establishes favorable
interactions with amino acid residues of the regions adjacent to the
active site of the target enzymes, such as the PTP1B secondary
noncatalytic pocket and the LMW-PTP variable loop, and, accord-
ingly, was shown to markedly influence both inhibitor potency and
selectivity of these compounds [50e53].
2. Results and discussion
2.1. Chemistry
(Z)-4-[(5-Arylidene-2,4-dioxothiazolidin-3-yl)methyl]benzoic
acids (5) were synthesized according to the general method
depicted in Scheme 1. The condensation of commercially available
2,4-thiazolidinedione with suitable aromatic aldehydes
3 in
refluxing ethanol in the presence of piperidine provided 5-
arylidene-2,4-thiazolidinediones 4. The reaction between com-
pounds 4 and 4-(bromomethyl)benzoic acid, using potassium
X
COOH
N
S
O
5
8
X = O
R
X = NPh
R'
a
b
c
d
e
R = H, R’ = OCH2CH2Ph
R = OCH2CH2Ph, R’ = H
R = OCH3, R’ = OCH2Ph
Therefore, compounds 1 and 2 appear to act as bidentate in-
hibitors of both PTPs that can target both the catalytic site and
adjacent loops. In compounds 2, additional hydrophobic in-
teractions between the 2-phenylimino ring and amino acid resi-
dues adjacent to the catalytic sites, such as Phe182 of PTP1B or
Tyr49 (IF1)/Trp49 (IF2) of LMW-PTP were also found to be
possible, in agreement with the greater inhibitory effectiveness of
several 2-phenylimino-4-thiazolidinone derivatives 2 compared
f
R = OCH2Ph, R’ = OCH3
R = OCH2Ph, R’ = F
g
R = OCH3, R’ = OCH2CH2Ph
R = OCH2CH2Ph, R’ = OCH3
Fig. 2. Structures of compounds 5aeg and 8aeg.

116
R. Ottanà et al. / European Journal of Medicinal Chemistry 71 (2014) 112e127
O
H
N
COOH
O
O
H
N
c, d
a, b
N
+
Ar CHO
S
O
S
S
O
O
3 a-g
R
R
R'
R'
4 a-g
5 a-g
Scheme 1. Reagents and conditions: (a) C5H11N, C2H5OH,
D; (b) CH3COOH; (c) BrCH2C6H4COOH, K2CO3,
D; (d) HCl.
carbonate as base in refluxing acetone, followed by a workup in
acidic medium and recrystallization from methanol, provided pure
acids 5aeg (Scheme 1). Although aldehydes 3aeg are commercially
available, they were synthesized as follows. 2-Phenylethoxy
substituted benzaldehydes 3aed were obtained by the reaction of
the suitable hydroxybenzaldehyde with (2-bromoethyl)benzene in
anhydrous DMF in the presence of potassium carbonate (Scheme
2). Benzyloxybenzaldehydes 3e and 3f were prepared by the
microwave-assisted O-alkylation of the appropriate hydrox-
ybenzaldehydes with benzyl chloride in anhydrous DMF in the
presence of sodium hydride (Scheme 3). For the synthesis of
aldehyde 3g, the reaction of O-alkylation between 3-fluoro-4-
hydroxybenzaldehyde and benzyl chloride was carried out in
refluxing DMF. 3-Fluoro-4-hydroxybenzaldehyde was prepared by
the demethylation of 3-fluoro-4-methoxybenzaldehyde with boron
tribromide in dichloromethane (Scheme 4).
The condensation of 4-[(4-oxo-2-phenyliminothiazolidin-3-yl)
methyl]benzoic acid (7) with aldehydes 3aeg, in refluxing ethanol
in the presence of piperidine as base, provided (Z)-4-[(5-arylidene-
4-oxo-2-phenyliminothiazolidin-3-yl)methyl]benzoic acids 8aeg
(Scheme 5).
Compound 7 was prepared by the condensation of 4-[(3-
phenylthioureido)methyl]benzoic acid (6) with chloroacetyl chlo-
ride in the presence of triethylamine as base in refluxing ethanol
[51]. As already reported, starting from N-aryl-N0-alkyl substituted
thioureas, derivative 7 is the only isomer obtained in these exper-
imental conditions [51e53,55]. The parent thiourea 6 was synthe-
sized by the reaction of phenylisothiocyanate and 4-(aminomethyl)
benzoic acid in refluxing ethanol [51].
CHO
CHO
a
OMe
+
PhCH2Cl
OMe
OCH2Ph
3 e, f
OH
Scheme 3. Reagents and conditions: (a) NaH, anhydrous DMF, MW.
characteristic triplets in the range from 3.03 ppm to 4.28 ppm,
whereas the benzyloxy methylene group of analogs 5eeg and 8eeg
resonated as a singlet at 5.07e5.26 ppm. In 13C NMR spectra, the
methylene carbon on N-3 resonated in the range between 44.3 ppm
and 45.9 ppm, whereas the methylene carbons of the ethoxy chain
gave rise to signals at 34.8e35.2 ppm and 68.4e69.4 ppm.
As observed for the previously synthesized analogs and
confirmed by means of X-ray crystallographic analysis [55,56], the
Knoevenagel condensation that provided compounds 4 and 8 gave
rise only to the thermodynamically stable Z isomers in every case.
In fact, the 1H NMR spectra of acids 5 and 8, as well as those of
precursors 4, showed only one singlet attributable to the resonance
of the 5-methylidene proton in the range from 7.60 ppm to
7.96 ppm. The carbon nucleus of the same methylidene group
resonated at 130.3e134.3 ppm, as resulted from 1He13C hetero-
correlation experiments.
2.2. Enzyme inhibition assays
Compounds 5aeg and 8aeg were tested to assess their inhibi-
tory activity toward human recombinant PTP1B, the two active
isoforms (IF1 and IF2) of human LMW-PTP and human TC-PTP
(Table 1).
The structures of all compounds were assigned on the basis of
their analytical and spectroscopic (1H and 13C NMR, MALDI-TOF
mass spectrometry) data.
Beyond the standard 1H and 13C NMR 1D spectra, the complete
and unambiguous assignment of signals was provided by several
experiments with 2D techniques, as 1H homocorrelated TOCSY and
ROESY and 1He13C hetero-correlated HSQC and HMBC. In the 1H
NMR spectra of benzoic acids 5 and 8, a diagnostic singlet attrib-
utable to the resonance of the methylene group on N-3 of the
thiazolidinone ring was present in the range from 4.78 ppm to
4.97 ppm (compounds 5) and from 5.12 ppm to 5.26 ppm (com-
pounds 8). In addition, the resonance of the two methylene groups
of the ethoxy chain of compounds 5aed and 8aed gave rise to two
All the tested benzoic acid derivatives 5aeg and 8aeg inhibited
PTP1B, with IC50 values lower than 10
of 5c (IC50 ¼ 24.6 M). Out of them, compounds 5b, 5d, 8b, 8c, 8d, 8f
and 8g exerted marked inhibitory effects, with IC50 values ranging
from 1.1 M to 2.8 M (Table 1).
mM, with the only exception
m
m
m
The enzyme inhibition results evidenced structureeactivity re-
lationships consistent with those already observed for parent
compounds 1, 2 [50e53]. In fact, 2-phenylimino-4-thiazolidinone
derivatives 8, with the exception of 8d and 8e, were shown to be
more potent PTP1B inhibitors than their 2,4-thiazolidinedione
CHO
CHO
CHO
CHO
CHO
a
b
a
R'
Br
+
Ph
F
F
F
O
R'
OMe
OH
HO
OCH2Ph
Ph
3 a-d
3 g
R' = H, OCH3
Scheme 4. Reagents and conditions: (a) BBr3, CH2Cl2, 0 ꢁC; (b) NaH, C6H5CH2Cl,
anhydrous DMF,
Scheme 2. Reagents and conditions: (a) K2CO3, anhydrous DMF, 40e50 ꢁC.
D.

R. Ottanà et al. / European Journal of Medicinal Chemistry 71 (2014) 112e127
117
S
a
COOH
NCS
+
COOH
N
N
b
H2N
6
c, d
N
S
COOH
N
S
COOH
N
N
O
O
7
R
R'
8 a-g
Scheme 5. Reagents and conditions: (a) C2H5OH,
D
; (b) ClCH2COCl, Et3N, C2H5OH, D; (c) ArCHO 3aeg, C5H11N, C2H5OH, D; (d) CH3COOH.
counterparts 5 bearing the same 5-arylidene moiety, thus con-
firming the favorable role of the 2-phenylimino moiety in
enhancing the affinity for the target enzyme. Moreover, out of the
2-phenylethoxy substituted compounds 5aed and 8aed, the de-
rivatives bearing the phenylethoxy group in the para position of the
5-benzylidene ring (5b, 5d, 8b and 8d) exhibited excellent PTP1B
inhibitory activity. Indeed, the displacement of the substituent to
the meta position resulted in decreased activity (compounds 5a, 5c
and 8a), with the only exception of 8c which displayed an activity
level moderately higher than that of its para-substituted isomer 8d
(Table 1). This is in agreement with the generally greater potency of
5-(4-phenoxybenzylidene) substituted compounds 1a, 2a and 5-
(4-benzyloxybenzylidene) substituted analogs 1b, 2b (Fig. 1) of the
series 1, 2 compared with their meta-substituted isomers [50e53].
Out of benzyloxybenzylidene substituted analogs (5e, 5f, 8e and
8f), compound 8f was shown to be the most potent PTP1B inhibitor,
resulting to be more effective than both homologous 8d and meta-
benzyloxy substituted isomer 8e (Table 1). Subsequently, the
methoxy group in the meta position of the benzylidene ring of 8f
was replaced with a more lipophilic and smaller fluorine substit-
uent, resulting in derivative 8g, which, although producing
appreciable enzyme inhibition, was slightly less active than parent
8f. Analogously, fluorosubstituted derivative 5g displayed an IC50
value higher than that of 5f (Table 1).
of their benzyloxy substituted counterparts 1b, 1c, 2b, 2c (Fig. 1)
[50,51], 5e, 5f, 8e and 8f (Fig. 2) highlighted that the elongation of
the chain connecting the two aromatic rings of the 5-arylidene
portion resulted in effective inhibitors of the target PTP1B,
although it did not generally provide a marked gain in potency.
Compounds 5d and 8c (IC50 ¼ 1.1
Table 1) were more active inhibitors than their benzyloxy analogs
5f and 8e (IC50 ¼ 4.6 M and 3.9 M, respectively), whereas only
mM and 1.7 mM, respectively,
m
m
compounds 5a and 5c, which are both 3-(2-phenylethoxy)benzy-
lidene substituted in position 5 of the thiazolidinedione ring,
resulted significantly less active than their respective benzylox-
ybenzylidene substituted counterparts 1c (Fig. 1) [50] and 5e.
Therefore, the 5-(2-phenylethoxy)benzylidene portion of com-
pounds 5aed and 8aed may critically contribute to the affinity of
inhibitors for the target enzyme, especially when the phenylethoxy
substituent is in the para position of the benzylidene ring, further
highlighting that the 5-arylidene moiety can exert a remarkable
influence on the PTP1B inhibitory activity of 4-[(5-arylidene-4-oxo-
2-arylimino/oxothiazolidin-3-yl)methyl]benzoic acids.
All tested benzoic acid-based PTP1B inhibitors 5 and 8, with the
exception of 5c, also proved to inhibit human LMW-PTP, without
reaching the inhibitory potency observed against PTP1B. In fact, we
observed from twice to more than 45-fold selectivity towards
PTP1B over LMW-PTP. Moreover, in agreement with the inhibition
profile of the previously investigated analogs [50e53], compounds
5 and 8 displayed from 2- to more than 5.5-fold selectivity towards
isoform IF1 over IF2 (Table 1). As already observed for the inhibition
On the whole, comparing the in vitro inhibitory profiles of
phenylethoxy substituted compounds 5aed and 8aed with those
of PTP1B, 2-phenylimino-4-thiazolidinones
8 were shown to
be more potent IF1 inhibitors than the corresponding 2,4-
thiazolidinediones 5, with the exception of 8d which was as
active as its counterpart 5d, and again compounds that are phe-
nylethoxy or benzyloxy substituted in the para position of the 5-
benzylidene ring were generally more effective inhibitors than
their meta-substituted isomers (Table 1).
Compounds 5 and 8 inhibited TC-PTP generally at low micro-
molar doses (Table 1). Compounds 5b, 5d, 5e, 8c and 8f showed a
2e3-fold selectivity toward PTP1B over the highly homologous TC-
PTP, whereas the other analogs inhibited both enzymes with
similar potency or were more effective against TC-PTP.
In order to establish the mechanism of PTP1B inhibition, we
performed additional tests on all compounds to verify if they were
reversible or irreversible inhibitors. Appropriate aliquots of PTP1B
were incubated in the presence of high molar excess of each in-
hibitor for 1 h at 37 ꢁC. Control experiments were performed
adding DMSO instead of inhibitor. Then, the enzyme mixtures were
diluted 400-fold with the assay buffer, and the residual enzyme
Table 1
IC50 values of compounds 5 and 8 against human PTP1B, IF1 and IF2 LMW-PTP
isoforms and TC-PTP.
Compd.
IC50
(
m
M)
PTP1B
IF1
IF2
TC-PTP
5a
5b
5c
5d
5e
5f
5g
8a
8b
8c
8d
8e
8f
9.5 ꢂ 0.3
2.6 ꢂ 0.2
24.6 ꢂ 1.8
1.1 ꢂ 0.2
4.0 ꢂ 0.1
4.6 ꢂ 0.2
7.5 ꢂ 0.3
4.6 ꢂ 0.2
1.5 ꢂ 0.3
1.7 ꢂ 0.2
2.8 ꢂ 0.2
3.9 ꢂ 0.1
1.6 ꢂ 0.1
2.2 ꢂ 0.1
20.1 ꢂ 0.4
20.1 ꢂ 0.5
>50
>50
>50
>50
>50
>50
>50
>50
26.8 ꢂ 0.8
8.9 ꢂ 0.2
>50
7.3 ꢂ 0.5
5.9 ꢂ 0.4
16.9 ꢂ 1.6
3.5 ꢂ 0.3
12.7 ꢂ 0.4
3.5 ꢂ 0.3
10.8 ꢂ 1.1
1.9 ꢂ 0.3
0.7 ꢂ 0.1
4.3 ꢂ 0.8
3.4 ꢂ 0.4
4.0 ꢂ 0.2
5.1 ꢂ 0.4
0.8 ꢂ 0.1
10.8 ꢂ 0.2
22.3 ꢂ 0.4
10.3 ꢂ 0.3
34.3 ꢂ 1.4
10.9 ꢂ 0.2
4.0 ꢂ 0.1
17.2 ꢂ 0.7
10.7 ꢂ 0.3
9.1 ꢂ 0.2
4.5 ꢂ 0.1
3.3 ꢂ 0.1
19.7 ꢂ 0.3
>50
10.0 ꢂ 0.2
15.5 ꢂ 0.2
8g

118
R. Ottanà et al. / European Journal of Medicinal Chemistry 71 (2014) 112e127
activity was assayed. The results shown in Fig. 3 demonstrated that
the recovery of activity for compounds 5aeg, 8c, 8d, and 8f was
quite satisfactory (80e90% range), indicating that they behaved as
reversible inhibitors. In contrast, in these experimental conditions,
the recovery of activity for compounds 8a, 8b, 8e and 8g was not
fully satisfactory (50e60% range), indicating that they behaved as
partially reversible inhibitors.
docking studies performed into the PTP1B active site (see below), it
can be inferred that the 2-phenylethoxy moiety may influence the
binding mode of inhibitors to the enzyme by modulating the
effectiveness of the interaction with the secondary noncatalytic
pocket.
2.4. Docking studies
Docking studies were performed to gain insight into the pro-
teineinhibitor interactions within the enzyme binding sites. PDB
[57] entry 2QBS [58] was selected as template for human PTP1B.
5PNT [39] and 1XWW [40] were selected as structures for human
LMW-PTP IF1 and IF2, respectively. The selected protein structures
were processed with CCDCs software GOLD version 5.1 for docking
experiments. After docking, each structure was minimized in the
binding site using the software package LigandScout (force field
MMFF94) [59] and prioritized using a 3D pharmacophore model
explaining the most plausible docking poses.
With PTP1B, almost all generated conformations for compounds
5 and 8 were oriented with the carboxylic acid towards the catalytic
center. The generally observed poses for compounds 5 involved
electrostatic interactions and hydrogen bonds between the car-
boxylic acid and the side chains of Lys120 and Arg221, as illustrated
in Fig. 5 for compound 5d docked into the enzymatic cavity of
PTP1B. The phenyl ring of the p-methylbenzoic acid residue of the
2.3. Kinetic studies
In order to determine the type of inhibition, we analyzed
experimental data by the double reciprocal plot method. For each
compound, we measured the initial hydrolysis rate using seven
substrate final concentrations (0.5e25 mM range), in the presence
of increasing concentrations of each compound. Compounds 5b, 5d,
8b, 8c and 8d, which are the most active PTP1B inhibitors among 5-
(2-phenylethoxy)benzylidene substituted compounds 5aed and
8aed, were selected for kinetic studies. They produced three
differing plot types as follows: i) compound 8c behaved as a
competitive inhibitor, since the graph showed straight lines that
intersect each other on the 1/v axis (Fig. 4D); ii) compound 5d
behaved as a noncompetitive inhibitor, since the graph showed
straight lines that intersect each other on the abscissa axis (Fig. 4B);
iii) finally, compounds 5b, 8b and 8d behaved as mixed-type in-
hibitors, since their graphs showed straight lines that intersect each
other in the left panel (Fig. 4A, C and E). The competitive inhibition
model shown in Scheme 6A indicates that the inhibitor binds only
to the free enzyme (E), and that the enzymeeinhibitor complex (EI)
is non-productive, while in the noncompetitive model shown in
Scheme 6B the inhibitor is capable to bind both E and the enzymee
substrate complex (ES), and the enzymeesubstrateeinhibitor
ternary complex (ESI) is non-productive. Finally, the mixed-type
inhibitor model shown in Scheme 6C indicates that the inhibitor
binds both E and ES. In this case, EI has lower affinity for S than E
ligand interacts by p-stacking with the side chain of Phe182, on one
side, and hydrophobic interaction with Ala217 and Tyr46 on the
other side. The side chain of Ile219 accommodates the 5-
benzylidene ring holding the R and R0 substituents, while the
distal phenyl ring is stabilized by a
Arg254 and hydrophobic interaction with Met258 in the secondary
noncatalytic pocket of the enzyme. similar orientation is
p-cation interaction with
A
observed with compounds of series 8, although they go deeper into
the binding pocket of PTP1B. Most common poses for compounds 8
are stabilized through electrostatic and H-bond interactions with
key residues in the catalytic center (Phe182, Gln266 and Arg221).
As observed in series 5, the aromatic moiety holding the carboxylic
(a
> 1), and ESI is non-productive.
For each compound we calculated the Ki value, using the
appropriate equations. Results obtained in this experiment are re-
ported in Table 2. We can observe that compounds 5b, 5d, 8b, and
8d, in which the 2-phenylethoxy group is present in the para po-
sition of the 5-benzylidene ring, behave as mixed-type or
noncompetitive inhibitors. Instead, compound 8c, which bears the
2-phenylethoxy group in the meta position of the 5-benzylidene
ring, inhibited the enzyme competitively. Accordingly to the
acid of 8 is stabilized by
pep stacking and hydrophobic contacts
with Phe182 (WPD loop), Tyr46 (YRD loop) and Ala217. The most
important difference between the two series is observed at the exit
of the binding cleft accommodating the benzoic acid moiety. There,
the side chain of Lys120 interacts and stabilizes the phenylimino
moiety in position 2 of the thiazolidinone cycle by an H-bond
interaction with the nitrogen, and by a
p-cation interaction with
the aromatic ring. On the opposite side, the distal phenyl ring of the
5-aryldiene moiety is accommodated in the noncatalytic pocket
and stabilized by a p-cation interaction with the positively charged
guanidine group of Arg24.
From the stabilizing interactions for compounds 5 and 8, which
are illustrated in Figs. 5 and 6 with the most potent PTP1B inhibitor
of each series 5d and 8b, respectively, two 3D pharmacophore
models were derived. The relative activities of the two series of
compounds against PTP1B can be discussed based on their ability to
form these interactions. For example, none of the 20 conformations
generated for 5c (data not shown) fulfilled all of the above-
mentioned pharmacophore features at the same time, justifying a
weaker affinity for PTP1B. Interestingly, the docking study showed
that 3-methoxy-4-phenylethoxy substitution on the benzylidene
ring leads to a good stabilization of the complex PTP1B/5d (Fig. 5).
However, the in silico analysis of series 5 highlighted that the in-
teractions with amino acid residues of the secondary noncatalytic
pocket cannot be maintained when the positions of the methoxy
and phenylethoxy moieties are switched on the benzylidene ring,
as in the case of 5c. Nevertheless, this does not explain the high
inhibitory potency of analogous compound 8c. However, the
Fig. 3. Inhibition reversibility assay of compounds 5 and 8. Control experiments were
carried out adding DMSO (C). Additional details are described in the Experimental
section. Data represent the mean ꢂ SE of three replicates.

R. Ottanà et al. / European Journal of Medicinal Chemistry 71 (2014) 112e127
119
Fig. 4. Double reciprocal plot (1/v versus 1/[S]). Compound 5b (A): 0
4.5 M, ; 5.5 M, A. Compound 8b (C): 0 M, C; 0.5 M, ,; 1.0 M, :; 1.5
M, B; 3.5 M, :; 4.5 M, ; 5.5
M, A. All tests were conducted in quadruplicate. The data represent the means ꢂ S.E.M.
m
M, -; 2.5
m
M, B; 3.5
m
M, :; 4.5
m
M, ; 5.5
m
m
M, A. Compound 5d (B): 0
M, -; 1.0 M, B; 1.5 M, :; 2.0 mM, . Compound 8d (E):
7
m
M, -; 2.5
m
M, B; 3.5
m
M, :;
7
m
m
m
m
m
m
M, B; 2.0
m
M, -. Compound 8c (D): 0
m
m
7
0
m
M, -; 2.5
m
m
m
m
7
Scheme 6. Inhibition model: A, competitive; B, noncompetitive; C, mixed. In the scheme: E, free enzyme; ES, enzymeesubstrate complex; EI, enzymeeinhibitor complex; ESI,
enzymeesubstrateeinhibitor ternary complex; P, reaction product.

120
R. Ottanà et al. / European Journal of Medicinal Chemistry 71 (2014) 112e127
Table 2
Ki values,
PTP1B.
8f and 8g are shown in Fig. 7. Interestingly, compounds lacking of
a
values and inhibition types of compounds 5b, 5d, 8b, 8c and 8d towards
the 2-phenylethoxy or benzyloxy group in the para position of the
5-benzylidene ring (such as 8a, 8c and 8e) or of phenylimino group
in position 2 of the thiazolidinone cycle (like all compounds in
series 5) generally showed lower activity for both IF isoenzymes,
highlighting the crucial role of hydrophobic contacts within the
enzymatic cavity.
Compd.
Ki (m
M)
a
Inhibition type
5b
5d
8b
8c
8d
2.3 ꢂ 0.5
1.4 ꢂ 0.2
0.5 ꢂ 0.1
1.5 ꢂ 0.1
1.5 ꢂ 0.4
1.8 ꢂ 0.5
e
Mixed
Noncompetitive
Mixed
Competitive
Mixed
3.0 ꢂ 0.2
e
2.0 ꢂ 0.2
2.5. Cellular assays
pharmacophore built for series 8 can explain this result. The ben-
zoic acid of the ligands can be seen as a central body of the molecule
deeply inserted into the catalytic center, with two arms starting
from the thiazolidinone ring: the phenylimino moiety on one side
(‘short arm’), and the benzylidene ring with its R and R0 sub-
stituents on the other side (‘long arm’). In an ideal enzymeeligand
interaction, both arms can establish stabilizing interactions within
the corresponding subsites. Since the phenylimino group (‘short
The cellular efficacy of compounds 5 and 8 was evaluated in
cultures of mouse C2C12 skeletal muscle cells, which constitutively
express PTP1B, TC-PTP and LMW-PTP as well as insulin receptor.
Tested compounds 5aeg and 8aeg did not interfere significantly
with the viability of these cells, which maintained ꢃ80% percentage
of their ability to reduce the MTT to formazan, with the exception of
5a (almost 60%) (Fig. 8).
In order to assess the capability of compounds 5aeg and 8aeg
to activate the insulin metabolic pathway, we evaluated the phos-
arm’) of compound 8c binds Lys120 by a
p-cation interaction, the
‘long arm’ shifts to the other side where the distal aromatic ring
reaches out to residue Arg24. Therefore, the larger arm span of
series 8 can explain their higher binding affinity to PTP1B.
phorylation levels of IR
b upon treatment of C2C12 cells with each
compound at 25 M concentration. Fig. 9 shows the effects pro-
m
duced by the tested compounds, compared to those elicited by the
control or by 10 nM insulin (a saturating dose for C2C12 cells [60]).
The smaller and less deep binding pocket of IF1 and IF2
compared to human PTP1B explains the lower activity of the tested
compounds towards these enzymes. Docking results consistently
show that the carboxylic acid of the most potent inhibitors of series
5 and 8 are trapped within the IF enzymatic cavity. In both isoforms,
relevant stabilizing interactions contain (i) hydrogen-bond donors
from the backbone nitrogen atoms of Leu13, Ile16, Cys17 and the
Cys12 sulfur, and (ii) the positively charged guanidinium group of
Arg18. The aromatic ring of the benzoic acid of the inhibitors forms
lipophilic contacts to the hydrophobic side chains of Leu13, Ile16
and Tyr131. Both enzymes stabilize the thiazolidinone moiety
either with Tyr49 (IF1) or Trp49 (IF2). Docking results show two
different orientations for short and long arm, respectively. When
interpreting the steric complementarity of the docking poses, a
tendency for the long arm to surround Leu13 can be observed for
IF1, while the long arm in IF2 points into the opposite direction
towards Ile16. The higher affinity for IF1 and IF2 of compounds 8
compared to series 5 indicated that the interactions involving the
phenylimino group significantly affect the inhibition profile of this
chemical scaffold. The opposite poses in IF1 and IF2 could be
explained by the different orientation of the phenylimino moiety to
Tyr49 (IF1) and Trp49 (IF2), respectively. Docking poses for 8b, 8d,
Most compounds induced increases of IR
levels higher than that of the control. Upon treatment with the
most active compounds (5c, 5e, and 8g) IR phosphorylation levels
were very close to that produced by 10 nM insulin suggesting that
they behave as insulinomimetic agents. Analogs 5d, 5f, 5g and 8be
d also induced an appreciable increase in the tyrosine phosphory-
lation of IRb compared to the control, without reaching the effect
produced by insulin (Fig. 9).
After incubation of compounds 5aeg and 8aeg in differentiated
C2C12 myotube cultures for 90 min, the cellular uptake of 2-
deoxyglucose (2-DOG) was also measured (Fig. 10).
Out of the tested compounds, 5b, 5c, 5eeg and 8bee were
shown to induce significant increases in the 2-DOG cellular uptake
that, interestingly, were comparable to that produced by 10 nM
insulin (Fig. 10). These findings demonstrated that these benzoic
acid derivatives are compounds capable to significantly improve
cellular insulin signaling, thus eliciting insulinomimetic effects
that culminate in a markedly enhanced glucose transport in the
cells.
b phosphorylation to
b
In our assay, compound 5a also produced an appreciable in-
crease in 2-DOG cellular incorporation, although with a lesser
Fig. 5. Main binding mode of 5d in PTP1B (PDB code 2QBS) in 3D (left) and 2D (right). The 3D pharmacophore is superposed to the ligand and illustrated with yellow spheres
indicating lipophilic areas, red arrows hydrogen bonds, and pink discs or -cation interactions. Variable size of amino acid labeling corresponds to the z-axis position (depth).
(For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
pep
p

R. Ottanà et al. / European Journal of Medicinal Chemistry 71 (2014) 112e127
121
Fig. 6. Main binding mode of 8b in PTP1B (PDB code 2QBS) in 3D (left) and 2D (right). The 3D pharmacophore is superposed to the ligand and illustrated with yellow spheres
indicating lipophilic areas, red arrows hydrogen bonds, and pink discs or -cation interactions. Variable size of amino acid labeling corresponds to the z-axis position (depth).
(For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
pep
p
extent. Analogs 5d and 8g showed a moderate activity, whereas 8a
and 8f were inactive in this assay (Fig. 10).
the introduction of the 2-phenylethoxy moiety was shown to in-
fluence not only the inhibitory effectiveness but also the inhibition
mechanism towards PTP1B. In fact, when the 2-phenylethoxy group
was present in the para position of the 5-benzylidene ring (com-
pounds 5b, 5d, 8b, 8d), a marked inhibitory effect was observed
resulting from a mixed-type or noncompetitive inhibition of PTP1B,
in contrast with the behavior of the previously assayed phenoxy and
benzyloxybenzylidene analogs of series 1, 2 which act as competi-
tive PTP1B inhibitors. Our docking studies supported the ration-
3. Conclusions
As a part of a project aimed to identify PTP inhibitors as agents
capable to fight insulin resistance, new 4-[(5-arylidene-4-oxo-
2-phenylimino/oxothiazolidin-3-yl)methyl]benzoic acids (5aeg,
8aeg) were synthesized and submitted to a series of in vitro assays
to evaluate their biological activity.
alization of these experimental results, also constituting
a
prospective rational basis for the development of optimized analogs.
Most of the new compounds of series 5 and 8 were shown to
inhibit human LMW-PTP and TC-PTP, however displaying a general
preference for human PTP1B. In particular, compounds 5b, 5d, 5e, 5g,
8c and 8f exhibited from moderate to significant selectivity toward
PTP1B over both TCPTP and LMW-PTP. In addition, most of the tested
compounds selectively inhibited isoform IF1 of human LMW-PTP
over isoenzyme IF2. The combined inhibition of these PTPs might
be beneficial for the effective treatment of a number of human dis-
eases whose pathogenetic course is strongly correlated to the
development of insulin and/or leptin resistance. Moreover, out of PTP
inhibitors 5, 8, we identified compounds endowed with good
insulinomimetic effects in C2C12 mouse skeletal muscle cells.
Systematic structural modifications aimed to optimize the activity
profile of 4-[(5-arylidene-4-oxothiazolidin-3-yl)methyl]benzoic acid
derivatives as PTP inhibitors allowed us to establish their major SARs.
Our docking studies and SAR analysis indicate that these compounds
can act as bidentate PTP inhibitors whose activity is strongly influ-
enced by the nature of the 5-arylidene portion and the substituent in
position 2 of the heterocyclic scaffold, while their p-methylbenzoic
acid moiety appears to be critical for the enzyme recognition.
In this context, the results of the evaluation of the new analogs of
series 5 and 8 further confirmed that the 5-arylidene moiety can act
as a leading structural motif which is able to exert significant effects
on the interaction mode of these inhibitors with the target enzymes.
In particular, the extension of the 5-arylidene portion by means of
Fig. 7. Proposed binding mode for compounds 8b, 8d, 8f and 8g superposed in IF1 (left, PDB code 5PNT) and in IF2 (right, PDB code 1XWW). The 3D pharmacophore is superposed
to the ligand, illustrated with yellow spheres indicating lipophilic areas and red arrows hydrogen bonds. Amino acid labels are scaled according to z-axis orientation (depth). (For
interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

122
R. Ottanà et al. / European Journal of Medicinal Chemistry 71 (2014) 112e127
Fig. 10. 2-Deoxyglucose uptake in C2C12 myotubes. Cells were treated with com-
pounds 5, 8 (25 M final concentration), or DMSO (C) (1.2% final concentration) for
90 min, or insulin (Ins, 10 nM) for 30 min. Additional details are described in the
Experimental section. Data represent the mean ꢂ SE of three replicates, and were
normalized versus protein concentration (*p < 0.05). The significativity of each datum
was calculated versus the control value.
Fig. 8. MTT responses in differentiated C2C12 cell cultures treated with compounds 5
and 8. (C), control value. Details are described in the Experimental section. Data
represent the mean ꢂ SE of three replicates.
m
Compounds 5b, 5c, 5eeg and 8bed were shown to be promising
agents, which were able to induce interesting increases in both IR
b
in a drug design programme aimed to the discovery of new drugs for
the treatment of insulin resistance-related pathologies.
phosphorylationand 2-DOG cellularuptake. Inparticular, theireffects
on glucose cellular uptake were remarkable, being very similar to that
produced by insulin. Therefore, along with benzoic acid-based ana-
logs that we recently identified, they can be proposed as lead com-
pounds of a class of insulinomimetic agents that could be developed
4. Experimental section
4.1. Chemistry
Melting points were recorded on a Kofler hot-stage apparatus
and are uncorrected. TLC controls were carried out on precoated
silica gel plates (F 254 Merck). Combustion analyses (C, H, N),
determined by means of a C. Erba mod. 1106 elem. Analyzer, were
within ꢂ0.4% of theory.
1H and 13C NMR spectra were achieved by a Bruker Avance
300 MHz and a Varian 500 MHz spectrometers equipped with
Linux workstations running TOPSPIN and vNMRj software pack-
ages respectively. Beyond the standard 1H and 13C NMR 1D analysis,
1H homocorrelated TOCSY and ROESY and 1He13C hetero-
correlation HSQC and HMBC experiments were performed. All
data were analyzed on a PC with the Mestrenova software package
(version 6.0.2, Mestrelab research S.L.). All the spectra were phased,
baseline was corrected where necessary and solvent signals (CDCl3
or DMSO-d6) were used as reference for both 1H and 13C spectra.
Matrix-assisted laser desorption/ionization time-of-flight
(MALDI-TOF) mass spectrometry was performed with a 4800
MALDI-TOF/TOF spectrometer (Applied Biosystems, Framingham,
MA, USA). The spectra were acquired in reflected mode. The MALDI
matrix was 2-[(2E)-3-(4-tert-butylphenyl)-2-methylprop-2-enyli
dene]malononitrile (DCTB).
A microwave oven Discover (CEM) was used to carry out
microwave-assisted reactions. Unless stated otherwise, all mate-
rials were obtained from commercial suppliers and used without
further purification. The purity of synthetic compounds was
established as ꢃ95% by combustion analysis.
General procedures for the synthesis of arylaldehydes 3aeg and
their chemical-physical and 1H NMR data are included in the
Supplementary material.
Fig. 9. IR
(Ins); ii) 1.2% DMSO (C); iii) 25
described in the Experimental section. Data represent the mean ꢂ SE of three repli-
cates (*p < 0.05). The significativity of each datum was calculated versus the control
value. Top, representative Western Blots; Bottom, data quantifications and statistical
analyses.
b
phosphorylation in C2C12 cells. Cells were treated with: i) 10 nM insulin
M compounds 5aeg and 8aeg. Additional details are
4.2. General procedure for the synthesis of 5-arylidene-2,4-
thiazolidinediones 4aeg
m
A mixture of 2,4-thiazolidinedione (2.34 g, 20 mmol), the
appropriate aldehyde 3 (20 mmol) and piperidine (1.36 g, 16 mmol)

R. Ottanà et al. / European Journal of Medicinal Chemistry 71 (2014) 112e127
123
in EtOH (150 ml) was refluxed for 24 h. The reaction mixture was
poured into water and acidified (pH 4e5) with AcOH to give
compounds 4 as solids which were recrystallized from methanol.
The solution was acidified (pH 3) with HCl and was stirred at room
temperature for 15 min. After evaporation to dryness in vacuo, the
crude solid was washed with water and recrystallized from meth-
anol to provide pure acids 5.
4.2.1. 5-{[3-(2-Phenylethoxy)phenyl]methylidene}-2,4-
thiazolidinedione (4a)
4.3.1. 4-[(5-{[3-(2-Phenylethoxy)phenyl]methylidene}-2,4-
Yield 41%; mp 145e148 ꢁC; 1H NMR (DMSO-d6):
d
3.06 (t,
dioxothiazolidin-3-yl)methyl]benzoic acid (5a)
J ¼ 6.9 Hz, 2H, CH2); 4.25 (t, J ¼ 6.9 Hz, 2H, OCH2); 7.06 (d, J ¼ 8.1 Hz,
1H, arom); 7.14e7.34 (m, 7H, arom); 7.44 (t, J ¼ 8.1 Hz, 1H, arom);
7.77 (s, 1H, CH methylidene). Anal. (C18H15NO3S) calcd: C 66.44, H
4.65, N 4.30; found: C 66.15, H 4.50, N 4.44.
Yield 67%; mp 186e188 ꢁC; 1H NMR (DMSO-d6):
d 3.04 (t,
J ¼ 6.5 Hz, 2H, CH2); 4.22 (t, J ¼ 6.5 Hz, 2H, OCH2); 4.89 (s, 2H,
NCH2); 7.05 (d, J ¼ 8.0 Hz, 1H, arom); 7.15 (m, 1H, arom); 7.17 (s, 1H,
arom); 7.21 (t, J ¼ 6.5 Hz,1H, arom); 7.30e7.32 (m, 4H, arom); 7.40e
7.43 (m, 3H, arom); 7.92 (s, 1H, CH methylidene); 7.93 (m, 2H,
4.2.2. 5-{[4-(2-Phenylethoxy)phenyl]methylidene}-2,4-
arom). 13C NMR (DMSO-d6):
d 34.8 (CH2); 44.4 (NCH2); 68.4 (OCH2);
thiazolidinedione (4b)
116.1, 117.3, 122.0, 126.4, 127.6, 128.3, 129.0, 129.7, 130.5 (CH arom);
133.6 (CH methylidene); 121.4 (5-C); 130.7, 134.3, 138.2,140.0,158.8
(Cq arom); 165.4, 167.1, 167.3 (C]O). MALDI-TOF MS (m/z): 458.13
(M ꢀ H)ꢀ. Anal. (C26H21NO5S) calcd: C 67.96, H 4.61, N 3.05; found:
C 68.17, H 4.80, N 2.92.
Yield 79%; mp 180e183 ꢁC; 1H NMR (CDCl3):
d
3.14 (t, J ¼ 6.6 Hz,
2H, CH2); 4.25 (t, J ¼ 6.6 Hz, 2H, OCH2); 7.00 (m, 2H, arom); 7.30e
7.35 (m, 5H, arom); 7.45 (m, 2H, arom), 7.81 (s,1H, CH methylidene).
Anal. (C18H15NO3S) calcd: C 66.44, H 4.65, N 4.30; found: C 66.70, H
4.87, N 4.51.
4.3.2. 4-[(5-{[4-(2-Phenylethoxy)phenyl]methylidene}-2,4-
4.2.3. 5-{[4-Methoxy-3-(2-phenylethoxy)phenyl]methylidene}-2,4-
dioxothiazolidin-3-yl)methyl]benzoic acid (5b)
thiazolidinedione (4c)
Yield 66%; mp 228e230 ꢁC; 1H NMR (DMSO-d6):
d 3.06 (t,
Yield 68%; mp 190 ꢁC; 1H NMR (CDCl3):
d
3.14 (t, J ¼ 7.5 Hz, 2H,
J ¼ 6.6 Hz, 2H, CH2); 4.28 (t, J ¼ 6.6 Hz, 2H, OCH2); 4.88 (s, 2H,
NCH2); 7.11 (m, 2H, arom); 7.22 (t, J ¼ 6.0 Hz, 1H, arom); 7.31e7.32
(m, 4H, arom); 7.39 (m, 2H, arom); 7.59 (m, 2H, arom), 7.91 (m, 2H,
CH2); 3.89 (s, 3H, OCH3); 4.22 (t, J ¼ 7.5 Hz, 2H, OCH2); 6.90e6.96
(m, 2H, arom); 7.08 (m, 1H, arom); 7.22e7.30 (m, 5H, arom); 7.67 (s,
1H, CH methylidene). Anal. (C19H17NO4S) calcd: C 64.21, H 4.82, N
3.94; found: C 64.03, H 5.00, N 3.79.
arom); 7.92 (s, 1H, CH methylidene). 13C NMR (DMSO-d6):
d 34.7
(CH2); 44.3 (NCH2); 68.5 (OCH2); 115.5, 126.3, 127.4, 128.3, 129.0,
129.6, 132.4 (CH arom); 117.7 (5-C); 133.6 (CH methylidene); 125.4,
131.4, 138.1, 139.7, 160.4 (Cq arom); 165.6, 167.1, 167.4 (C]O).
MALDI-TOF MS (m/z): 458.11 (M ꢀ H)ꢀ. Anal. (C26H21NO5S) calcd: C
67.96, H 4.61, N 3.05; found: C 68.18, H 4.79, N 3.11.
4.2.4. 5-{[3-Methoxy-4-(2-phenylethoxy)phenyl]methylidene}-2,4-
thiazolidinedione (4d)
Yield 71%; mp 185e188 ꢁC; 1H NMR (CDCl3):
d
3.21 (t, J ¼ 7.2 Hz,
2H, CH2); 3.94 (s, 3H, OCH3); 4.30 (t, J ¼ 7.2 Hz, 2H, OCH2); 6.95e
7.12 (m, 3H, arom); 7.28e7.35 (m, 5H, arom); 7.81 (s, 1H, CH
methylidene); 8.45 (br.s, 1H, NH). Anal. (C19H17NO4S) calcd: C 64.21,
H 4.82, N 3.94; found: C 64.44, H 4.69, N 4.09.
4.3.3. 4-[(5-{[4-Methoxy-3-(2-phenylethoxy)phenyl]methylidene}-
2,4-dioxothiazolidin-3-yl)methyl]benzoic acid (5c)
Yield 63%; mp 205e208 ꢁC; 1H NMR (DMSO-d6):
d 3.03 (t,
J ¼ 6.9 Hz, 2H, CH2); 3.79 (s, 3H, OCH3); 4.19 (t, J ¼ 6.9 Hz, 2H,
OCH2); 4.86 (s, 2H, NCH2); 7.09 (d, J ¼ 8.0 Hz, 1H, arom); 7.17 (d,
J ¼ 8.0 Hz, 1H, arom); 7.18 (s, 1H, arom); 7.20 (t, J ¼ 6.5 Hz, 1H,
arom); 7.29e7.31 (m, 4H, arom); 7.38 (m, 2H, arom); 7.85 (s, 1H, CH
4.2.5. 5-[(3-Benzyloxy-4-methoxyphenyl)methylidene]-2,4-
thiazolidinedione (4e)
Yield 27%; mp 210e211 ꢁC; 1H NMR (DMSO-d6):
d 3.80 (s, 3H,
OCH3); 5.16 (s, 2H, OCH2); 7.16e7.21 (m, 3H, arom); 7.35e7.43 (m,
5H, arom); 7.74 (s, 1H, CH methylidene). Anal. (C18H15NO4S) calcd: C
63.33, H 4.43, N 4.10; found: C 63.11, H 4.28, N 4.31.
methylidene); 7.90 (m, 2H, arom). 13C NMR (DMSO-d6):
d 35.1
(CH2); 44.6 (NCH2); 56.1 (OCH3); 69.3 (OCH2); 112.7, 115.1, 124.5,
126.7, 127.8, 128.6, 129.3, 130.0 (CH arom); 134.3 (CH methylidene);
118.1 (5-C); 125.8, 130.7, 138.4, 140.5, 148.3, 151.6 (Cq arom); 165.8,
167.4, 167.8 (C]O). MALDI-TOF MS (m/z): 488.10 (M ꢀ H)ꢀ. Anal.
(C27H23NO6S) calcd: C 66.24, H 4.74, N 2.86; found: C 66.40, H 5.02,
N 2.67.
4.2.6. 5-[(4-Benzyloxy-3-methoxyphenyl)methylidene]-2,4-
thiazolidinedione (4f)
Yield 44%; mp 183e185 ꢁC; 1H NMR (CDCl3):
d 3.96 (s, 3H,
OCH3); 5.22 (s, 2H, OCH2); 6.97e7.44 (m, 8H, arom); 7.73 (s, 1H, CH
methylidene); 8.30 (br. s, 1H, NH). Anal. (C18H15NO4S) calcd: C
63.33, H 4.43, N 4.10; found: C 63.59, H 4.61, N 3.87.
4.3.4. 4-[(5-{[3-Methoxy-4-(2-phenylethoxy)phenyl]methylidene}-
2,4-dioxothiazolidin-3-yl)methyl]benzoic acid (5d)
Yield 41%; mp 238e240 ꢁC; 1H NMR (DMSO-d6):
d 3.05 (t,
4.2.7. 5-[(4-Benzyloxy-3-fluorophenyl)methylidene]-2,4-
thiazolidinedione (4g)
J ¼ 7.0 Hz, 2H, CH2); 3.79 (s, 3H, OCH3); 4.24 (t, J ¼ 7.0 Hz, 2H,
OCH2); 4.88 (s, 2H, NCH2); 7.16e7.17 (m, 2H, arom); 7.20 (s, 1H,
arom); 7.21 (m, 1H, arom); 7.30e7.32 (m, 4H, arom); 7.39 (m, 2H,
arom); 7.90 (m, 2H, arom); 7.91 (s, 1H, CH methylidene). 13C NMR
Yield 25%; mp 199e202 ꢁC; 1H NMR (DMSO-d6):
d 5.28 (s, 2H,
OCH2); 7.35e7.48 (m, 8H, arom); 7.75 (s, 1H, CH methylidene). Anal.
(C17H12FNO3S) calcd: C 62.00, H 3.67, N 4.25; found: C 61.89, H 3.82,
N 4.06.
(DMSO-d6):
d 34.8 (CH2); 44.3 (NCH2); 55.7 (OCH3); 68.9 (OCH2);
113.2, 113.9, 124.0, 126.3, 127.5, 128.3, 129.0, 129.7 (CH arom); 133.9
(CH methylidene); 117.9 (5-C); 125.6,130.5,138.0,140.2,149.1, 150.3
(Cq arom); 165.5, 167.1, 167.4 (C]O). MALDI-TOF MS (m/z): 488.11
(M ꢀ H)ꢀ. Anal. (C27H23NO6S) calcd: C 66.24, H 4.74, N 2.86; found:
C 66.41, H 4.93, N 3.00.
4.3. General procedure for the synthesis of 4-[(5-arylidene-2,4-
dioxothiazolidin-3-yl)methyl]benzoic acids 5aeg
A mixture of the appropriate 5-arylidene-2,4-thiazolidinedione
4 (1.5 mmol), 4-(bromomethyl)benzoic acid (0.65 g, 3 mmol) and
potassium carbonate (0.83 g, 6 mmol) in anhydrous acetone
(150 ml) was refluxed for 48 h. The solvent was evaporated under
reduced pressure and the solid residue was dissolved in methanol.
4.3.5. 4-({5-[(3-Benzyloxy-4-methoxyphenyl)methylidene]-2,4-
dioxothiazolidin-3-yl}methyl)benzoic acid (5e)
Yield 79%; mp 248e250 ꢁC; 1H NMR (DMSO-d6):
d
3.84 (s, 3H,
OCH3), 4.89 (s, 2H, NCH2), 5.15 (s, 2H, OCH2); 7.15 (d, J ¼ 8.5 Hz, 1H,

124
R. Ottanà et al. / European Journal of Medicinal Chemistry 71 (2014) 112e127
arom); 7.24 (d, J ¼ 8.5 Hz, 1H, arom); 7.26 (s, 1H, arom); 7.33 (t,
J ¼ 7.0 Hz, 1H, arom); 7.39 (m, 2H, arom); 7.40 (d, J ¼ 6.5 Hz, 2H,
arom); 7.45 (m, 2H, arom); 7.87 (s, 1H, CH methylidene); 7.91 (m,
NCH2); 6.97 (d, J ¼ 6.5 Hz, 2H, arom); 7.01 (d, J ¼ 8.0 Hz, 2H, arom);
7.18 (m, 1H, arom); 7.19 (m, 1H, arom); 7.28 (m, 2H, arom); 7.29 (m,
2H, arom); 7.38 (t, J ¼ 6.5 Hz, 2H, arom); 7.44 (d, J ¼ 8.5 Hz, 2H,
arom); 7.47 (m, 2H, arom); 7.71 (s, 1H, CH methylidene); 7.92 (m,
2H, arom); 13C NMR (DMSO-d6):
d 44.4 (NCH2); 55.9 (OCH3); 70.0
(OCH2); 112.5, 115.0, 124.7, 127.6, 127.8, 128.1, 128.6, 129.7 (CH
arom); 133.9 (CH methylidene); 118.1 (5-C); 125.5, 130.5, 136.7,
140.3, 147.9, 151.6 (Cq arom); 165.6, 167.1, 167.5 (C]O). MALDI-TOF
MS (m/z): 474.06 (M ꢀ H)ꢀ. Anal. (C26H21NO6S) calcd: C 65.67, H
4.45, N 2.95; found: C 65.88, H 4.60, N 2.78.
2H, arom). 13C NMR (DMSO-d6):
d 35.0 (CH2); 45.8 (NCH2); 68.7
(OCH2); 115.7, 121.2, 123.9, 126.7, 127.5, 128.0, 128.9, 129.8, 130.0,
132.2 (CH arom); 131.2 (CH methylidene); 118.0 (5-C); 125.9, 130.7,
138.4, 141.1, 147.9, 160.5 (Cq arom); 150.1 (C]N); 166.3, 167.6 (C]
O). MALDI-TOF MS (m/z): 533.13 (M ꢀ H)ꢀ. Anal. (C32H26N2O4S)
calcd: C 71.89, H 4.90, N 5.24; found: C 72.11, H 5.28, N 4.96.
4.3.6. 4-({5-[(4-Benzyloxy-3-methoxyphenyl)methylidene]-2,4-
dioxothiazolidin-3-yl}methyl)benzoic acid (5f)
4.4.3. 4-[(5-{[4-Methoxy-3-(2-phenylethoxy)phenyl]methylidene}-
Yield 93%; mp 223e225 ꢁC; 1H NMR (DMSO-d6):
d
3.82 (s, 3H,
4-oxo-2-phenyliminothiazolidin-3-yl)methyl]benzoic acid (8c)
OCH3); 4.78 (s, 2H, NCH2); 5.07 (s, 2H, OCH2); 6.87 (m, 1H, arom),
Yield 48%; mp 232e235 ꢁC; 1H NMR (DMSO-d6):
d 3.00 (t,
7.01 (m, 1H, arom); 7.26e7.33 (m, 8H, arom); 7.67 (s, 1H, CH
J ¼ 7.0 Hz, 2H, CH2); 3.77 (s, 3H, OCH3); 4.14 (t, J ¼ 7.0 Hz, 2H, OCH2);
5.14 (s, 2H, NCH2); 6.96 (d, J ¼ 7 Hz, 2H, arom); 7.07 (m, 2H, arom);
7.17 (m, 1H, arom); 7.19 (s, 1H, arom); 7.21 (m, 1H, arom); 7.28 (m,
2H, arom); 7.36 (t, J ¼ 7.5 Hz, 2H, arom); 7.48 (m, 2H, arom); 7.73 (s,
1H, CH methylidene); 7.93 (m, 2H, arom). 13C NMR (DMSO-d6):
methylidene); 7.86 (m, 2H, arom); 13C NMR (DMSO-d6):
d 45.3
(NCH2); 56.1 (OCH3); 70.4 (OCH2); 114.1, 114.2, 124.3, 128.0, 128.4,
129.0, 130.2 (CH arom); 134.4 (CH methylidene); 118.6 (5-C); 126.3,
130.7, 137.0, 140.8, 149.7, 150.6 (Cq arom); 166.0, 167.9 (C]O).
MALDI-TOF MS (m/z): 474.04 (M ꢀ H)ꢀ. Anal. (C26H21NO6S) calcd: C
65.67, H 4.45, N 2.95; found: C 65.58, H 4.33, N 3.09.
d
35.0 (CH2); 45.6 (NCH2); 55.9 (OCH3); 69.3 (OCH2); 112.6, 115.5,
121.7, 123.2, 125.3, 126.5, 127.7, 128.5, 129.1, 129.6, 129.8 (CH arom);
131.4 (CH methylidene); 118.2 (5-C); 126.2, 130.3, 138.2, 141.1, 147.7,
148.1, 151.0 (Cq arom); 150.0 (C]N); 166.0, 167.3 (C]O). MALDI-
TOF MS (m/z): 563.33 (M ꢀ H)ꢀ. Anal. (C33H28N2O5S) calcd: C
70.19, H 5.00, N 4.96; found: C 70.30, H 5.23, N 4.81.
4.3.7. 4-({5-[(4-Benzyloxy-3-fluorophenyl)methylidene]-2,4-
dioxothiazolidin-3-yl}methyl)benzoic acid (5g)
Yield 62%; mp 297e300 ꢁC; 1H NMR (DMSO-d6):
d
4.89 (s, 2H,
NCH2); 5.25 (s, 2H, OCH2), 7.27e7.56 (m, 10H, arom); 7.88 (m, 2H,
arom); 7.90 (s, 1H, CH methylidene). 13C NMR (DMSO-d6):
44.5
d
4.4.4. 4-[(5-{[3-Methoxy-4-(2-phenylethoxy)phenyl]methylidene}-
(NCH2); 70.4 (OCH2); 115.7 (JCF ¼ 28.7 Hz), 127.6, 127.9, 128.0, 128.2,
128.4, 128.6, 128.7 (CH arom); 129.7 (CH methylidene); 119.8 (5-C);
126.2, 129.1, 130.9, 136.2 (JCF ¼ 45.0 Hz), 140.0, 148.2, 151.6
(JCF ¼ 243.7 Hz) (Cq arom); 165.5, 167.3 (C]O). MALDI-TOF MS (m/
z): 462.0 (M ꢀ H)ꢀ. Anal. (C25H18FNO5S) calcd: C 64.79, H 3.91, N
3.02; found: C 64.61, H 3.80, N 3.18.
4-oxo-2-phenyliminothiazolidin-3-yl)methyl]benzoic acid (8d)
Yield 42%; mp 232e234 ꢁC; 1H NMR (DMSO-d6):
d 3.00 (t,
J ¼ 7.0 Hz, 2H, CH2); 3.75 (s, 3H, OCH3); 4.20 (t, J ¼ 7.0 Hz, 2H,
OCH2); 5.13 (s, 2H, NCH2); 6.97 (d, J ¼ 7.5 Hz, 2H, arom); 7.02 (d,
J ¼ 8.5 Hz, 1H, arom); 7.18 (m, 1H, arom); 7.20 (m, 1H, arom); 7.28
(m, 2H, arom); 7.29 (m, 2H, arom); 7.38 (t, J ¼ 7.5 Hz, 2H, arom); 7.44
(m, 2H, arom); 7.47 (m, 2H, arom); 7.71 (s, 1H, CH methylidene);
4.4. General procedure for the synthesis of 4-[(5-arylidene-4-oxo-
7.92 (m, 2H, arom). 13C NMR (DMSO-d6):
d 35.0 (CH2); 45.8 (NCH2);
2-phenyliminothiazolidin-3-yl)methyl]benzoic acids 8aeg
55.7 (OCH3); 68.7 (OCH2); 115.7, 121.2, 122.5, 125.3, 126.7, 127.5,
128.6, 129.2, 129.8, 129.9, 132.2 (CH arom); 131.2 (CH methylidene);
118.0 (5-C); 125.9, 130.7, 138.4, 141.1, 142.0, 147.9, 149.0 (Cq arom);
150.1 (C]N); 166.3, 167.6 (C]O). MALDI-TOF MS (m/z): 563.11
(M ꢀ H)ꢀ. Anal. (C33H28N2O5S) calcd: C 70.19, H 5.00, N 4.96; found:
C 70.05, H 4.83, N 5.06.
A mixture of 4-[(4-oxo-2-phenyliminothiazolidin-3-yl)methyl]
benzoic acid (7) (0.49 g, 1.5 mmol), which was synthesized ac-
cording to a previously reported procedure [51], the appropriate
aldehyde 3 (1.5 mmol) and piperidine (0.25 g, 3 mmol) in EtOH
(150 ml) was refluxed for 72e96 h. The reaction mixture was
poured into water and acidified (pH 4e5) with glacial AcOH to give
compounds 8 as solids which were recrystallized from methanol.
4.4.5. 4-{[5-[(3-Benzyloxy-4-methoxyphenyl)methylidene]-4-oxo-
2-phenyliminothiazolidin-3-yl]methyl}benzoic acid (8e)
Yield 50%; mp 231e233 ꢁC; 1H NMR (DMSO-d6):
d 3.76 (s, 3H,
4.4.1. 4-[(4-Oxo-5-{[3-(2-phenylethoxy)phenyl]methylidene}-2-
OCH3), 5.12 (s, 2H, NCH2); 5.17 (s, 2H, OCH2); 6.94e7.52 (m, 15H,
phenyliminothiazolidin-3-yl)methyl]benzoic acid (8a)
arom); 7.78 (s, 1H, CH methylidene); 7.95 (m, 2H, arom). 13C NMR
Yield 66%; mp 127e130 ꢁC; 1H NMR (DMSO-d6):
d
2.98 (t,
(DMSO-d6): d 45.9 (NCH2); 56.3 (OCH3); 70.4 (OCH2); 114.1, 115.0,
J ¼ 7.0 Hz, 2H, CH2); 4.08 (t, J ¼ 7.0 Hz, 2H, OCH2); 5.12 (s, 2H, NCH2);
6.82 (d, J ¼ 8.0 Hz, 1H, arom); 6.88 (m, 2H, arom); 6.94 (d, J ¼ 7.5 Hz,
1H, arom); 7.09 (t, J ¼ 7.5 Hz,1H, arom); 7.13 (t, J ¼ 7.5 Hz,1H, arom);
7.17 (m, 2H, arom); 7.20 (m, 2H, arom); 7.23 (m, 1H, arom); 7.28 (t,
J ¼ 7.5 Hz, 2H, arom); 7.46 (m, 2H, arom); 7.57 (m,1H, arom); 7.62 (s,
1H, CH methylidene); 7.91 (m, 2H, arom). 13C NMR (DMSO-d6):
121.5, 122.8,125.3, 128.0,128.3,128.4,128.8, 129.9,130.0 (CH arom);
131.6 (CH methylidene); 118.6 (5-C); 125.9, 126.8, 136.9,141.3, 147.8,
149.9, 153.2 (Cq arom); 149.7 (C]N); 166.2, 167.4 (C]O). MALDI-
TOF MS (m/z): 549.18. (M ꢀ H)ꢀ. Anal. (C32H26N2O5S) calcd: C
69.80, H 4.76, N 5.09; found: C 70.02, H 4.98, N 4.95.
d
34.9 (CH2); 45.4 (NCH2); 68.3 (OCH2); 115.6, 115.8, 120.9, 121.6,
4.4.6. 4-{[5-[(4-Benzyloxy-3-methoxyphenyl)methylidene]-4-oxo-
124.8, 126.0, 127.8, 128.0, 128.3, 128.5, 129.5, 129.7 (CH arom); 130.3
(CH methylidene); 121.2 (5-C); 130.1, 134.4, 137.5, 140.1, 147.2, 158.6
(Cq arom); 149.2 (C]N); 165.8, 167.4 (C]O). MALDI-TOF MS (m/z):
533.11 (M ꢀ H)ꢀ. Anal. (C32H26N2O4S) calcd: C 71.89, H 4.90, N 5.24;
found: C 72.05, H 5.12, N 5.01.
2-phenyliminothiazolidin-3-yl]methyl}benzoic acid (8f)
Yield 64%; mp 212e215 ꢁC; H NMR (DMSO-d6):
d 3.74 (s, 3H,
OCH3); 5.10 (s, 2H, NCH2); 5.14 (s, 2H, OCH2); 6.97 (d, J ¼ 8.0 Hz, 2H,
arom); 7.01 (d, J ¼ 8.5 Hz, 1H, arom); 7.14 (d, J ¼ 8.5 Hz, 1H, arom);
7.18 (m, 1H, arom); 7.19 (s, 1H, arom), 7.31 (t, J ¼ 7.0 Hz, 1H, arom);
7.37 (m, 2H, arom); 7.38 (m, 2H, arom); 7.40 (m, 2H, arom); 7.48 (m,
2H, arom); 7.73 (s, 1H, CH methylidene); 7.93 (m, 2H, arom). 13C
4.4.2. 4-[(4-Oxo-5-{[4-(2-phenylethoxy)phenyl]methylidene}-2-
phenyliminothiazolidin-3-yl)ethyl]benzoic acid (8b)
NMR (DMSO-d6): d 45.7 (NCH2); 56.0 (OCH3); 70.2 (OCH2); 113.9,
114.8, 121.3, 122.6, 125.3, 127.8, 128.2, 128.4, 128.8, 129.8, 129.9 (CH
arom); 131.5 (CH methylidene); 118.5 (5-C); 126.5, 130.3, 136.7,
Yield 70%; mp 238e241 ꢁC; 1H NMR (DMSO-d6):
d
3.00 (t,
J ¼ 6.6 Hz, 2H, CH2); 4.20 (t, J ¼ 6.6 Hz, 2H, OCH2); 5.13 (s, 2H,

R. Ottanà et al. / European Journal of Medicinal Chemistry 71 (2014) 112e127
125
141.2, 147.9, 149.4, 150.0 (Cq arom); 149.8 (C]N); 166.2, 167.5 (C]
O). MALDI-TOF MS (m/z): 549.16 (M ꢀ H)ꢀ. Anal. (C32H26N2O5S)
calcd: C 69.80, H 4.76, N 5.09; found: C 69.96, H 4.61, N 5.18.
initial rates at different substrate concentrations. Experimental
data were analyzed using the Michaelis equation and a non-linear
fitting program FigSys [66]. Inhibition constants were determined
measuring initial hydrolysis rates at differing substrate and inhib-
itor concentrations. The apparent Km values measured at the
various inhibitor concentrations were plotted against inhibitor
concentration to calculate the Ki values. All measurements of initial
rate were carried out in triplicate. For each inhibitor, IC50 was
determined by measuring the initial hydrolysis rate under fixed p-
nitrophenylphosphate concentration, equal to the Km value of each
tested PTP. Data were fitted to the following equation using the
FigSys program:
4.4.7. 4-{[5-[(4-Benzyloxy-3-fluorophenyl)methylidene]-4-oxo-2-
phenyliminothiazolidin-3-yl]methyl}benzoic acid (8g)
Yield 30%; mp 262e265 ꢁC; 1H NMR (DMSO-d6):
d 5.16 (s, 2H,
NCH2); 5.22 (s, 2H, OCH2); 6.99 (m, 2H, arom); 7.19 (t, J ¼ 7.0 Hz, 1H,
arom); 7.29e7.49 (m, 12H, arom); 7.72 (s, 1H, CH methylidene); 7.93
(m, 2H, arom). 13C NMR (DMSO-d6):
d 45.7 (NCH2); 70.5 (OCH2);
115.9, 121.2, 124.5, 125.2, 126.8, 127.6, 127.8, 127.9, 128.1, 128.5, 128.8
(CH arom); 129.7 (CH methylidene); 118.3 (5-C); 126.6, 133.4, 134.3,
136.2, 147.7, 147.9, 156.4 (Cq arom); 149.5 (C]N); 165.9, 167.4 (C]
O). MALDI-TOF MS (m/z): 537.25 (M ꢀ H)ꢀ. Anal. (C31H23FN2O4S)
calcd: C 69.13, H 4.30, N 5.20; found: C 69.24, H 4.19, N 5.03.
Max ꢀ Min
y ¼
þ Min
1 þ ðx=IC50Þslope
where y ¼ vi/vo, i.e. the ratio between the activity measured in the
presence of the inhibitor (vi) and the activity of the control without
the inhibitor (vo). The parameter “x” is the inhibitor concentration.
To verify if compounds 5aeg and 8aeg were reversible in-
hibitors, appropriate aliquots of PTP1B were incubated in the
presence of at least 25-fold molar excess of inhibitor for 1 h at 37 ꢁC.
Control experiments were performed adding DMSO instead of in-
hibitor. After this interval time, the enzyme solutions were diluted
400-fold and the residual enzyme activity was assayed.
4.5. Docking studies
Protein and ligand preparations were carried out using the
software tool MOE [61] and CORINA [62]. Docking studies were
performed with GOLD [63] version 5.1 using default parameters
(GOLDScore, 100% search efficiency). PDB entries 2QBS [58] (human
PTP1B), 5PNT [39] (LMW-PTP IF1) and 1XWW [40] (LMW-PTP IF2)
were selected as protein templates. The active site was determined
ꢀ
by selecting all residues with any atom within 7 A from the outside
atoms the co-crystallized ligand for 2QBS, and the analog residues
for the other two PDB structures using the sequence alignment and
structure overlay functionality of the software tool MOE. Docking
was performed without any constraints and all compounds were
minimized using LigandScout’s [59] general purpose MMFF94
implementation after docking. In order to get the most probable
4.8. Cellular assays
C2C12 cells were obtained from the European Collection of Cell
Cultures (Salisbury, UK). Assays were carried out using differenti-
ating mouse C2C12 myoblasts which constitutively express both
PTP1B and IR. Cells were grown in DMEM, supplemented with 10%
binding poses,
a 3D pharmacophore was developed with
LigandScout, using the docking poses of the most active ligands
synthesized in this work. These 3D pharmacophores were then
utilized to select the docking poses that formed the basis for the
analysis described. LigandScout was also used for visualization,
analysis and illustrations.
FBS, 2 mM L-glutamine, 100 U/ml penicillin, and 100 mg/ml strep-
tomycin at 37 ꢁC in 5% CO2 humidified atmosphere. To induce dif-
ferentiation, the medium of 90% sub-confluent cells was replaced
by DMEM supplemented with 2% horse serum, and cells were
incubated at 37 ꢁC. After 2 days, the medium was replaced by fresh
differentiation medium, and cells were incubated for additional 2
days at the same conditions. Before experiments, the obtained
myotubes were starved for 20 h in the above medium without
4.6. Enzyme section
The complete coding sequences of human IF1, IF2 LMW-PTPs
and human PTP1B were cloned in frame with the sequence of the
glutathione S-transferase in the pGEX-2T bacterial expression
vector. Enzyme expression and purification were achieved in the
Escherichia coli TB1 strain [64]. Briefly, the recombinant fusion
proteins were purified from bacterial lysate using a single step af-
finity chromatography on glutathione-Sepharose. The solution
containing purified fusion proteins was treated with thrombin for
3 h at 37 ꢁC. Then the enzymes were purified from GST and
thrombin by gel filtration on Superdex G-75. The purity of protein
preparations was analyzed by SDS-polyacrylamide gel electropho-
resis according to Laemmli [65]. The T-cell PTP was purchased from
New England Biolabs Inc.
horse serum. Compounds 5aeg and 8aeg (25 mM final concentra-
tion) or DMSO (1.2% final concentration) were added to the culture
medium and the cells were incubated for 90 min at 37 ꢁC. The
positive control with insulin (10 nM final concentration) was car-
ried out incubating cells for 30 min a 37 ꢁC. Cells were then lysed
with SDS Laemmli sample buffer and analyzed by SDS-PAGE and
Western Blot using both anti-phosphorylated insulin receptor and
anti-actin antibodies (Santa Cruz Biotechnology Inc.) [67]. Quanti-
tative analyses of Western Blots were carried out using the KODAK
1D Image Analysis software.
To perform the cytotoxicity assays, DMEM was supplemented
with MTT (0.5 mg/ml final concentration) and then myotubes were
incubated at 37 ꢁC. After 2 h, the medium was removed and cells
were lysed with DMSO. The absorption levels of formazan were
read at 570 nm.
4.7. Phosphatase assay and inhibition experiments
To determine glucose uptake, C2C12 cells were seeded in 24-
well plates, differentiated as described above, and serum-starved
for 20 h. Then the myotubes were treated with compounds 5aeg
Phosphatase assay was carried out at 37 ꢁC using p-nitro-
phenylphosphate as substrate and the final volume was 1 ml. The
assay buffer (pH 7.0) contained 0.075 M
b
,b
-dimethylglutarate
and 8aeg (25 mM final concentration) or 1.2% DMSO for 90 min. The
buffer, 1 mM EDTA and 5 mM dithiothreitol. The reactions were
initiated by addition of aliquots of the enzyme preparations and
stopped at appropriate times by adding 4 ml of 1 M KOH. The
released p-nitrophenolate ion was determined by reading the
absorbance at 400 nm (ε ¼ 18,000 Mꢀ1 cmꢀ1). The main kinetic
parameters (Km and Vmax) were determined by measuring the
positive control with insulin (10 nM final concentration) was car-
ried out incubating cells for 30 min a 37 ꢁC.
Glucose transport was assayed as follows: after inhibitor or
other treatments, 0.1
m
Ci of 2-deoxy-D-1,2-[3H]-glucose (20 Ci/
mmol) in Ringer-Hepes buffer was added to each well and glucose
uptake was allowed at 37 ꢁC for 15 min. Cells were subsequently

126
R. Ottanà et al. / European Journal of Medicinal Chemistry 71 (2014) 112e127
sterol regulatory element-binding protein-1 gene expression, J. Biol. Chem.
278 (2003) 43095e43101.
[20] R.J. Gum, L.L. Gaede, S.L. Koterski, M. Heindel, J.E. Clampit, B.A. Zinker,
J.M. Trevillyan, R.G. Ulrich, M.R. Jirousek, C.M. Rondinone, Reduction of pro-
tein tyrosine phosphatase 1B increases insulin-dependent signalling in ob/ob
mice, Diabetes 52 (2003) 21e28.
washed with cold PBS and lysed in 0.2 M NaOH. Aliquots of the
lysates were counted with a -scintillation-counter, and data were
normalized against protein concentration, which was measured by
Bradford method [68].
b
[21] M. Delibegovic, K.K. Bence, N. Mody, E.G. Hong, H.J. Ko, J.K. Kim, B.B. Kahn,
B.G. Neel, Improved glucose homeostasis in mice with muscle-specific dele-
tion of protein-tyrosine phosphatase 1B, Mol. Cell. Biol. 27 (2007) 7727e7734.
[22] M. Delibegovic, D. Zimmer, C. Kauffman, K. Rak, E.G. Hong, Y.R. Cho, J.K. Kim,
B.B. Kahn, B.G. Neel, K.K. Bence, Liver-specific deletion of protein-tyrosine
phosphatase 1B (PTP1B) improves metabolic syndrome and attenuates diet-
induced endoplasmic reticulum stress, Diabetes 58 (2009) 590e599.
[23] I. Nieto-Vazquez, S. Fernandez-Veledo, C. de Alvaro, C.M. Rondinone,
A.M. Valverde, M. Lorenzo, Protein-tyrosine phosphatase 1B-deficient myo-
cytes show increased insulin sensitivity and protection against tumor necrosis
Acknowledgments
Work supported in part by the ‘Fondazione Cassa di Risparmio di
Pistoia e Pescia’, and by University of Firenze and University of
Messina.
Appendix A. Supplementary data
factor-a-induced insulin resistance, Diabetes 56 (2007) 404e413.
[24] K.K. Bence, M. Delibegovic, B. Xue, C.Z. Gorgun, G.S. Hotamisligil, B.G. Neel,
B.B. Kahn, Neuronal PTP1B regulates body weight, adiposity and leptin action,
Nat. Med. 12 (2006) 917e924.
[25] J.G. Mielke, C. Taghibiglou, L. Liu, Y. Zhang, Z. Jia, K. Adeli, Y.T. Wang, a
biochemical and functional characterization of diet-induced brain insulin
resistance, J. Neurochem. 93 (2005) 1568e1578.
[26] G. Liu, Technology evaluation: ISIS-113715, Isis, Curr. Opin. Mol. Ther. 6
(2004) 331e336.
[27] S. Zhang, Z.Y. Zhang, PTP1B as a drug target: recent developments in PTP1B
inhibitor discovery, Drug Discov. Today 12 (2007) 373e381.
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.ejmech.2013.11.001.
References
[1] (a) International Diabetes Federation, IDF Diabetes Atlas, 2012 Update, fifth
ed., International Diabetes Federation, Brussels, Belgium, 2011. http://www.
idf.org/diabetesatlas;
[28] A.P. Combs, Recent advances in the discovery of competitive protein tyrosine
phosphatase 1B inhibitors for the treatment of diabetes, obesity and cancer,
J. Med. Chem. 53 (2010) 2333e2344.
[29] D. Popov, Novel protein tyrosine phosphatase 1B inhibitors: interaction re-
quirements for improved intracellular efficacy in type 2 diabetes mellitus and
obesity control, Biochem. Biophys. Res. Commun. 410 (2011) 377e381.
[30] S. Thareja, S. Aggarwal, T.R. Bhardwaj, M. Kumar, Protein tyrosine phospha-
tase 1B inhibitors: a molecular level legitimate approach for the management
of diabetes mellitus, Med. Res. Rev. 32 (2012) 459e517.
(b) S. Jain, S. Saraf, Type 2 diabetes mellitus e its global prevalence and
therapeutic strategies, Diabetes Metab. Syndr. Clin. Res. Rev. 4 (2010) 48e56.
[2] D.J. Rader, Effect of insulin resistance, dyslipidemia, and intra-abdominal
adiposity on the development of cardiovascular disease and diabetes melli-
tus, Am. J. Med. 120 (2007) S12eS18.
[3] D.M. Muoio, C.B. Newgard, Molecular and metabolic mechanisms of insulin
resistance and
b-cells failure in type 2 diabetes, Nat. Rev. Mol. Cell Biol. 9
(2008) 193e205.
[4] R. Meshkani, K. Adeli, Hepatic insulin resistance, metabolic syndrome and
[31] K.A. Lantz, S.G. Emeigh Hart, S.L. Planey, M.F. Roitman, I.A. Ruiz-White,
H.R. Wolfe, M.P. McLane, Inhibition of PTP1B by trodusquemine (MSI-1436)
causes fat-specific weight loss in diet-induced obese mice, Obesity 18 (2010)
1516e1523.
cardiovascular disease, Clin. Biochem. 42 (2009) 1331e1346.
[5] L.F. Del Aguila, K.P. Claffey, J.P. Kirwan, TNF-a impairs insulin signaling and
insulin stimulation of glucose uptake in C2C12 muscle cells, Am. J. Physiol.
Endocrinol. Metab. 276 (1999) E849eE855.
[32] G. Scapin, S.B. Patel, J.W. Becker, Q. Wang, C. Desponts, D. Waddleton, K. Skorey,
W. Cromlish, C. Bayly, M. Therien, J.Y. Gauthier, C.S. Li, C.K. Lau, C. Ramachandran,
B.P. Kennedy, E. Asante-Appiah, The structural basis for the selectivity of ben-
zotriazole inhibitors of PTP1B, Biochemistry 42 (2003) 11451e11459.
[33] L. Bialy, H. Waldmann, Inhibitors of protein tyrosine phosphatases: next-
generation drugs? Angew. Chem. Int. Ed. 44 (2005) 3814e3839.
[34] R. Maccari, R. Ottanà, Low molecular weight phosphotyrosine protein phos-
phatases as emerging targets for the design of novel therapeutic agents,
J. Med. Chem. 55 (2012) 2e22.
[35] Y.A. Puius, Y. Zhao, M. Sullivan, D.S. Lawrence, S.C. Almo, Z.Y. Zhang, Identi-
fication of a second aryl phosphate-binding site in protein-tyrosine phos-
phatase 1B: a paradigm for inhibitor design, Proc. Natl. Acad. Sci. U. S. A. 94
(1997) 13420e13425.
[6] A. Cheng, N. Uetani, P.D. Simoncic, V.P. Chaubey, A. Lee-Loy, J. McGlade,
B.P. Kennedy, M.L. Tremblay, Attenuation of leptin action and regulation of
obesity by protein tyrosine phosphatase 1B, Dev. Cell 2 (2002) 497e503.
[7] D.L. Morris, L. Rui, Recent advances in understanding leptin signaling and
leptin resistance, Am. J. Physiol. Endocrinol. Metab. 297 (2009) E1247eE1259.
[8] E.C. McNay, C.T. Ong, R.J. McCrimmon, J. Cresswell, J.S. Bogan, R.S. Sherwin,
Hippocampal memory processes are modulated by insulin and high-fat-
induced insulin resistance, Neurobiol. Learn. Mem. 93 (2010) 546e553.
[9] D. Bosco, A. Fava, M. Plastino, T. Montalcini, A. Pujia, Possible implications of
insulin resistance and glucose metabolism in Alzheimer’s disease pathogen-
esis, J. Cell. Mol. Med. 15 (2011) 1807e1821.
[10] A. Apostolatos, S. Song, S. Acosta, M. Peart, J.E. Watson, P. Bickford,
D.R. Cooper, N.A. Patel, Insulin promotes neuronal survival via the alterna-
[36] A. Cheng, N. Dubè, F. Gu, M.L. Tremblay, Coordinated action of protein tyrosine
phosphatases in insulin signal transduction, Eur. J. Biochem. 269 (2002)
1050e1059.
tively spliced protein kinase C delta II (PKCdII) isoform, J. Biol. Chem. 287
(2012) 9299e9310.
[11] S.M. de la Monte, Insulin resistance and Alzheimer’s disease, BMB Rep. 42
(2009) 475e481.
[12] A.R. Saltiel, C.R. Kahn, Insulin signalling and the regulation of glucose and lipid
metabolism, Nature 414 (2001) 799e806.
[37] P. Chiarugi, P. Cirri, F. Marra, G. Raugei, G. Camici, G. Manao, G. Ramponi,
LMW-PTP is a negative regulator of insulin-mediated mitotic and metabolic
signalling, Biochem. Biophys. Res. Commun. 238 (1997) 676e682.
[38] S.K. Pandey, X.X. Yu, L.M. Watts, M.D. Michael, K.W. Sloop, A.R. Rivard,
T.A. Leedom, V.P. Manchem, L. Samadzadeh, R.A. McKay, B.P. Monia, S. Bhanot,
Reduction of low molecular weight protein-tyrosine phosphatase expression
improves hyperglycemia and insulin sensitivity in obese mice, J. Biol. Chem.
282 (2007) 14291e14299.
[13] S. Schinner, W.A. Scherbaum, S.R. Bornstein, A. Barthel, Molecular mecha-
nisms of insulin resistance, Diabetic Med. 22 (2005) 674e682.
[14] M. Elchebly, P. Payette, E. Michaliszyn, W. Cromlish, S. Collins, A.L. Loy,
D. Normandin, A. Cheng, J. Himms-Hagen, C.C. Chan, C. Ramachandran,
M.J. Gresser, M.L. Tremblay, B.P. Kennedy, Increased insulin sensitivity and
obesity resistance in mice lacking the protein tyrosine phosphatase-1B gene,
Science 283 (1999) 1544e1548.
[15] L.D. Klaman, O. Boss, O.D. Peroni, J.K. Kim, J.L. Martino, J.M. Zabolotny,
N. Moghal, M. Lubkin, Y.B. Kim, A.H. Sharpe, A. Stricker-Krongrad,
G.I. Shulman, B.G. Neel, B.B. Kahn, Increased energy expenditure, decreased
adiposity, and tissue-specific insulin sensitivity in protein-tyrosine phospha-
tase 1B-deficient mice, Mol. Cell. Biol. 20 (2000) 5479e5489.
[39] M. Zhang, C.V. Stauffacher, D. Lin, R.L. Van Etten, Crystal structure of a human
low molecular weight phosphotyrosyl phosphatase. Implications for substrate
specificity, J. Biol. Chem. 273 (1998) 21714e21720.
[40] A.P.R. Zabell, A.D. Schroff Jr., B.E. Bain, R.L. Van Etten, O. Wiest,
C.V. Stauffacher, Crystal structure of the human b-form low molecular weight
ꢀ
phosphotyrosyl phosphatase at 1.6-A resolution, J. Biol. Chem. 281 (2006)
6520e6527.
[41] S. Galic, M. Klingler-Hoffmann, M.T. Fodero-Tavoletti, M.A. Puryer, T.C. Meng,
N.K. Tonks, T. Tiganis, Regulation of insulin receptor signaling by the protein
tyrosine phosphatase TCPTP, Mol. Cell. Biol. 23 (2003) 2096e2108.
[42] L.F. Iversen, K.B. Moller, A.K. Pedersen, G.H. Peters, A.S. Petersen,
H.S. Andersen, S. Branner, S.B. Mortensen, N.P.H. Moller, Structure determi-
[16] R. Di Paola, L. Frittitta, G. Miscio, M. Bozzali, R. Baratta, M. Centra,
D. Spampinato, M.G. Santagati, T. Ercolino, C. Cisternino, T. Soccio,
S. Mastroianno, V. Tassi, P. Almgren, A. Pizzuti, R. Vigneri, V.A. Trischitta,
Variation in 30 UTR of hPTP1B increases specific gene expression and associ-
ates with insulin resistance, Am. J. Hum. Genet. 70 (2002) 806e812.
[17] J.L. Bento, N.D. Palmer, J.C. Mychaleckyj, L.A. Lange, C.D. Langefeld, S.S. Rich,
B.I. Freedman, D.W. Bowden, Association of protein tyrosine phosphatase 1B
gene polymorphisms with type 2 diabetes, Diabetes 53 (2004) 3007e3012.
[18] J.M. Zabolotny, K.K. Bence-Hanulec, A. Stricker-Krongrad, F. Haj, Y. Wang,
Y. Minokoshi, Y.B. Kim, J.K. Elmquist, L.A. Tartaglia, B.B. Kahn, B.G. Neel, PTP1B
regulates leptin signal transduction in vivo, Dev. Cell 2 (2002) 489e495.
[19] S. Shimizu, S. Ugi, H. Maegawa, K. Egawa, Y. Nishio, T. Yoshizaki, K. Shi,
Y. Nagai, K. Morino, K. Nemoto, T. Nakamura, M. Bryer-Ash, A. Kashiwagi,
Protein-tyrosine phosphatase 1B as new activator for hepatic lipogenesis via
nation of
T cell protein-tyrosine phosphatase, J. Biol. Chem. 277 (2002)
19982e19990.
[43] S. Galic, C. Hauser, B.B. Kahn, F.G. Haj, B.G. Neel, N.K. Tonks, T. Tiganis, Co-
ordinated regulation of insulin signaling by the protein tyrosine phosphatases
PTP1B and TCPTP, Mol. Cell. Biol. 25 (2005) 819e829.
[44] A. Fukushima, K. Loh, S. Galic, B. Fam, B. Shields, F. Wiede, M.L. Tremblay,
M.J. Watt, S. Andrikopoulos, T. Tiganis, T-cell protein tyrosine phosphatase
attenuates STAT3 and insulin signaling in the liver to regulate gluconeogen-
esis, Diabetes 59 (2010) 1906e1914.

R. Ottanà et al. / European Journal of Medicinal Chemistry 71 (2014) 112e127
127
[45] K. Loh, T.L. Merry, S. Galic, B.J. Wu, M.J. Watt, S. Zhang, Z.Y. Zhang, B.G. Nee,
T. Tiganis, T cell protein tyrosine phosphatase (TCPTP) deficiency in muscle
does not alter insulin signaling and glucose homeostasis in mice, Diabetologia
55 (2012) 468e478.
[46] K. Loh, A. Fukushima, X. Zhang, S. Galic, D. Briggs, P.J. Enriori, S. Simonds,
F. Wiede, A. Reichenbach, C. Hauser, N.A. Sims, K.K. Bence, S. Zhang,
Z.Y. Zhang, B.B. Kahn, B.G. Neel, Z.B. Andrews, M.A. Cowley, T. Tiganis,
Elevated hypothalamic TCPTP in obesity contributes to cellular leptin resis-
tance, Cell Metab. 14 (2011) 684e699.
[57] H.M. Berman, J. Westbrook, Z. Feng, G. Gilliland, T.N. Bhat, H. Weissig,
I.N. Shindyalov, P.E. Bourne, The protein data bank, Nucleic Acids Res. 28
(2000) 235e242.
[58] D.P. Wilson, Z.K. Wan, W.X. Xu, S.J. Kirincich, B.C. Follows, D. Joseph-McCarthy,
K. Foreman, A. Moretto, J. Wu, M. Zhu, E. Binnun, Y.L. Zhang, M. Tam, D.V. Erbe,
J. Tobin, X. Xu, L. Leung, A. Shilling, S.Y. Tam, T.S. Mansour, J. Lee, Structure-
based optimization of protein tyrosine phosphatase 1B inhibitors: from the
active site to the second phosphotyrosine binding site, J. Med. Chem. 50
(2007) 4681e4698.
[47] J. Montalibet, B.P. Kennedy, Therapeutic strategies for targeting PTP1B in
diabetes, Drug Discov. Today Ther. Strat. 2 (2005) 129e135.
[48] A.J. Nichols, R.D. Mashal, B. Balkan, Toward the discovery of small molecule
PTP1B inhibitors for the treatment of metabolic diseases, Drug Dev. Res. 67
(2006) 559e566.
[59] (a) G. Wolber, T. Langer, LigandScout: 3-D pharmacophores derived from
protein-bound ligands and their use as virtual screening filters, J. Chem. Inf.
Model. 45 (2004) 160e169;
(b) G. Wolber, T. Seidel, F. Bendix, T. Langer, Molecule-pharmacophore
superpositioning and pattern matching in computational drug design, Drug
Discov. Today 13 (2008) 23e29;
(c) G. Wolber, A. Dornhofer, T. Langer, Efficient overlay of small organic
molecules using 3D pharmacophores, J. Comput. Aided Mol. Des. 20 (2006)
773e788.
[49] K.M. Heinonen, A. Bourdeau, K.M. Doody, M.L. Tremblay, Protein tyrosine
phosphatases PTP-1B and TC-PTP play nonredundant roles in macrophage
development and IFN-
9372.
g signaling, Proc. Natl. Acad. Sci. 106 (2009) 9368e
[50] R. Maccari, P. Paoli, R. Ottanà, M. Jacomelli, R. Ciurleo, G. Manao, T. Steindl,
T. Langer, M.G. Vigorita, G. Camici, 5-Arylidene-2,4-thiazolidinediones as in-
hibitors of protein tyrosine phosphatases, Bioorg. Med. Chem. 15 (2007)
5137e5149.
[51] R. Ottanà, R. Maccari, R. Ciurleo, P. Paoli, M. Jacomelli, G. Manao, G. Camici,
C. Laggner, T. Langer, 5-Arylidene-2-phenylimino-4-thiazolidinones as PTP1B
and LMW-PTP Inhibitors, Bioorg. Med. Chem. 17 (2009) 1928e1937.
[52] R. Maccari, R. Ottanà, R. Ciurleo, P. Paoli, G. Manao, G. Camici, C. Laggner,
T. Langer, Structure-based optimization of benzoic acids as inhibitors of
protein tyrosine phosphatase 1B and low molecular weight protein tyrosine
phosphatase, ChemMedChem 4 (2009) 957e962.
[53] R. Ottanà, R. Maccari, S. Amuso, G. Wolber, D. Schuster, S. Herdlinger,
G. Manao, G. Camici, P. Paoli, New 4-[(5-arylidene-2-arylimino-4-oxo-3-
thiazolidinyl)methyl]benzoic acids active as protein tyrosine phosphatase
inhibitors endowed with insulinomimetic effect on mouse C2C12 skeletal
muscle cells, Eur. J. Med. Chem. 50 (2012) 332e343.
[60] P. Cirri, M.L. Taddei, P. Chiarugi, F. Buricchi, A. Caselli, P. Paoli, E. Giannoni,
G. Camici, G. Manao, G. Raugei, G. Ramponi, Insulin inhibits platelet-derived
growth factor-induced cell proliferation, Mol. Biol. Cell 16 (2005) 73e83.
[61] Molecular Operating Environment (MOE), 2011.10, Chemical Computing
Group Inc., 1010 Sherbooke St. West, Suite #910, Montreal, QC, Canada, H3A
2R7, 2011.
[62] J. Sadowski, J. Gasteiger, G. Klebe, Comparison of automatic three-dimensional
model builders using 639 X-ray structures, J. Chem. Inf. Comp. Sci. 34 (1994)
1000e1008.
[63] G. Jones, P. Willett, R.C. Glen, A.R. Leach, R. Taylor, Development and valida-
tion of a genetic algorithm for flexible docking, J. Mol. Biol. 267 (1997) 727e
748.
[64] R. Marzocchini, M. Bucciantini, M. Stefani, N. Taddei, M.G.M.M. Thunnissen,
P. Nordlund, G. Ramponi, Expression, purification and preliminary crystal
analysis of the human low Mr phosphotyrosine protein phosphatase isoform
1, FEBS Lett. 426 (1998) 52e56.
[54] U. Iannaccone, A. Bergamaschi, A. Magrini, G. Marino, N. Bottini, P. Lucarelli,
E. Bottini, F. Gloria-Bottini, Serum glucose concentration and ACP1 genotype
in healthy adult subjects, Metab. Clin. Exp. 54 (2005) 891e894.
[55] R. Ottanà, R. Maccari, M.L. Barreca, G. Bruno, A. Rotondo, A. Rossi, G. Chiricosta,
R. Di Paola, L. Sautebin, S. Cuzzocrea, M.G. Vigorita, 5-Arylidene-2-imino-4-
thiazolidinones: design and synthesis of novel antiinflammatory agents,
Bioorg. Med. Chem. 13 (2005) 4243e4252.
[65] U.K. Laemmli, cleavage of structural proteins during the assembly of the head
of bacteriophage T4, Nature 227 (1970) 680e685.
[66] FigSys, A Graphing, Charting and Data Analysis Software Package, BIOSOFTÒ
2D Dolphin Way, Stapleford Cambridge, CB2 5DW, UK, 2003.
,
[67] P. Paoli, P. Cirri, A. Caselli, F. Ranaldi, G. Bruschi, A. Santi, G. Camici, The
insulin-mimetic effect of morin: a promising molecule in diabetes treatment,
Biochim. Biophys. Acta 1830 (2013) 3102e3111.
[56] G. Bruno, L. Costantino, C. Curinga, R. Maccari, F. Monforte, F. Nicolò, R. Ottanà,
M.G. Vigorita, Synthesis and aldose reductase inhibitory activity of 5-
arylidene-2,4-thiazolidinediones, Bioorg. Med. Chem. 10 (2002) 1077e1084.
[68] M.M. Bradford, Rapid and sensitive method for the quantitation of microgram
quantities of protein utilizing the principle of protein-dye binding, Anal.
Biochem. 72 (1976) 248e254.